The optical performance of the microscope objective lens directly affects the quality of microscopic imaging, so it is of great significance to detect its wavefront aberration. A method for detecting the aberration of a microscopic objective lens based on a Hartmann wavefront sensor was proposed. This method employed the two-sphere method to calibrate systematic errors, enabling aberration characterization of microscope objectives. The model was established and simulated and analysed, and the results show that measured wavefront aberrations closely matched actual values, verifying model validity. The experimental setup was constructed, and the aberration detection of the microscope objective was carried out. The measurement shows the objective lens's root mean square (RMS) detection accuracy is ≤ 10 nm, with a repeatability accuracy of < 0.3 nm. This confirms the practical feasibility of the detection model. The method is simple and easy to set up, providing an detection model for wavefront aberration detection of medium and low power finite-conjugate microscope objectives.

Addressing the challenge of unclear segmentation boundaries arising from multi-component aliasing in coke optical tissue images, this paper proposes a MD-UNet semantic segmentation model. This model employs VGG16 as its backbone network and incorporates the CloAttention module at the deepest level of the encoder. By leveraging context-aware local enhancement and a global attention mechanism, CloAttention enables the model to focus better on critical image regions and enhances the perception of the complex textures inherent in coke optical tissues. Furthermore, a multi-branch dilated fusion (MBDF) module has been designed to replace the conventional convolution modules in the decoder. This substitution aims to effectively preserve and integrate multi-scale information, thereby enriching feature representation and mitigating information loss and detail blurring. Finally, the GELU activation function is adopted in place of ReLU to address the vanishing gradient problem encountered during network training. Comparative experiments on semantic segmentation models demonstrate that the proposed MD-UNet model achieves the most superior segmentation performance on coke optical tissues, reaching mIoU and F1-Score values of 88.72% and 94.28%, respectively. These results significantly outperform traditional semantic segmentation models, thereby validating the effectiveness of MD-UNet in enhancing the segmentation accuracy of coke optical tissues.

In recent years, research on the dynamic regulation of electromagnetic wave absorbers using graphene has attracted extensive attention. In this paper, a patterned graphene absorber is designed with a catenary, and the calculated results show that the absorber achieves perfect dual-frequency absorption at 5.57 THz and 7.11 THz in the terahertz frequency band in the TE wave mode. The analysis of the electric field strength and current distribution on the surface of the graphene layer shows that the perfect absorption of the dual frequency is related to the electric dipole resonance caused by the local surface plasmon excitation of the graphene layer. On this basis, the simulation of the relationship between the resonant absorption frequency and the geometric parameters of the absorber structure, the coupled bias voltage and the incident angle shows that the absorber can effectively realize the dynamic control of the absorption, and the structure can also reflect the absorption rate of more than 90% for a large range of incidence angle (0° to 40°).

When measuring the front surfaces of transparent objects by using phase measuring deflectometry (PMD), the reflections from the front and back surfaces of the object are superimposed, resulting in parasitic reflection, which makes it difficult to accurately reconstruct the front surface of objects. In this paper, we proposed a front-surface PMD method that suppresses parasitic reflections of transparent objects. Firstly, the initial phase of the front surface was extracted by continuous wavelet transform. Then, an optimized model was constructed by combining the multi-frequency phase-shifting method to obtain the accurate phase. Finally, the gradient integral was used to restore the three-dimensional (3D) morphology of the transparent object surface. The surfaces of transparent glasses and plano-convex lenses were measured using the above theory. Compared with the multi-frequency phase-shifting method, the reconstruction error of the 3 mm glass plate decreased from 21.81 μm to 15.72 μm, the error of the 4 mm glass plate is reduced from 19.98 μm to 13.46 μm, and the radius of curvature error of the plano-convex lens is reduced from 39.44 μm to 16.59 μm, thereby improving the accuracy of the front surface of transparent objects measurements.

Bone tissue is a heterogeneous material with a complex structure. Large cracks are prone to propagate into the bone prolonging postoperative recovery. A pulsed laser is used to cut the bone tissue, and a heterogeneous structure bone model related to the direction of the osteon is established. The stress intensity factors in the directions transverse, parallel, and across to the osteon are 2.70, 2.03, and 1.94 MPa·m0.5, respectively. The surface roughness in the directions transverse, parallel, and across to the osteon are 10.14, 7.12, and 6.98 μm, respectively. The stress intensity factor and roughness of the laser surface cut in the direction transverse to the osteons are higher compared with the parallel and across directions. The value of stress intensity factor, surface roughness and the crack propagation patterns are similar when the laser cutting direction is parallel and across to the osteons. Results show that the laser cutting only needs to consider two characteristic directions, i.e., the directions perpendicular and parallel to the osteon. The roughness of laser-cut surfaces in the directions, perpendicular, parallel, and across to the osteon is lower than that of mechanical-cut surface, indicating that laser cutting is more conducive to postoperative bone healing.

A measurement method for mapping the reflectivity distribution of curved high-reflectivity mirrors based on optical feedback cavity ring-down (OF-CRD) technique is introduced. The setup consists of a high-stability folded optical cavity and a high-precision positioning system with 5 degrees of freedom (DOF). By mathematical modeling, we established a universal mathematical relationship between the curved surface described by mathematical formula and the required adjustment for sample positioning, enabling dynamic five-dimensional compensation for optical elements with varying curvatures, apertures, and surface profiles. Through high-precision automated scanning strategy, the high-reflectivity distributions of samples with different surface profiles, apertures, and curvature diameters are obtained, this measurement is achieved by two-dimensionally raster-scanning while compensating for the other three dimensions (depth, pitch, and yaw) of the samples using a control program to ensure strict normal alignment between the probe beam and local surface throughout the measurement process. A reflectivity measurement repeatability error at the part-per-million (ppm) level is experimentally achieved. Good agreement is observed between the CRD-measured reflectivity profile and the grayscale image obtained by a laser confocal microscopy. Compared with the reflectivity values obtained by manual high-precision alignment, both reflectivity distributions obtained via automatic and manual alignments are consistent, demonstrating the reliability of the reflectivity distribution obtained with the experimental apparatus.

Optical fingerprint liveness detection plays a crucial role in preventing spoofing attacks on fingerprint recognition systems. Existing deep learning-based methods require large amounts of labeled data, while fingerprint image acquisition remains challenging. A few-shot optical fingerprint liveness detection method, featuring the fusion of spatial and frequency domain features, has been proposed for few-shot scenarios. Performance in liveness detection under few-shot conditions is enhanced using a bidirectional cross-domain attention mechanism and a high-frequency enhancement factor. Experimental results demonstrate outstanding performance on two benchmark datasets. With only 10 samples, the average classification error rate (ACER) reaches as low as 0.21% and 0.45%, outperforming existing methods. Notably, excellent adaptability in cross-sensor detection is shown. The method expands the practical applications of fingerprint liveness detection technology.

To better utilize the optical fiber bandwidth and enhance the communication capability of the RoF system, this paper combines photonic vector modulation and optical millimeter-wave generation techniques, proposing an 18-fold frequency 16 quadrature amplitude modulation (QAM) millimeter-wave generation scheme based on polarization division multiplexing (PDM). The system consists of two parallel modules: frequency doubling and vector modulation. In the frequency doubling module, two dual-parallel Mach-Zehnder modulators (DP-MZM) and an optical phase shifter are used to generate the ±6th-order optical sidebands. After passing through a semiconductor optical amplifier (SOA), four-wave mixing effects are induced, and the 18th-order sidebands are obtained through filtering. In the vector modulation module, eight 10 Gbit/s binary NRZ signals drive the DP-MZM, and polarization multiplexing technology is applied to implement 16QAM optical domain modulation on two orthogonal polarization directions. The PDM-16QAM signal is coupled with the 18th-order optical sidebands and, after photodetection, an 18-fold frequency PDM-16QAM signal is generated. Simulation analysis shows that when the photodetector's received power is ?25 dBm, the signal-to-noise ratio (SNR) of the system is greater than 20.5 dB, and after 30 km transmission through single-mode optical fiber, the system performance remains good. Even when there are fluctuations in the SOA injection current and laser linewidth within a certain range, the bit error rate (BER) of the system remains below 3.8×10?3.

A single-station real-time multi-target 3D coordinate measurement system is proposed to address the challenges in traditional multi-base station positioning systems that rely on multi-source observation intersection mechanisms. Such systems are often hindered by occlusions in complex environments and face difficulties in station layout adjustments. Achieving single-station real-time multi-target 3D sensing and measurement requires a measurement module with parallel range angle sensing capability to decouple the strong correlation between range and angle measurements, thus meeting the demands of real-time automatic multi-target measurement. This paper presents a novel method combining ultra-wideband ranging and rotational laser scanning for angle measurement. A comprehensive calibration model and measurement model for the range angle fusion system are established, and cooperative targets are designed with an analysis of the effect of target decentering errors. Additionally, a compensation method for systematic errors in the UWB ranging system is studied. Finally, a prototype system is developed for feasibility verification. Experimental results demonstrate that the proposed single-station measurement system achieves an average point error of less than 30 mm, representing nearly a 70% reduction in error compared to traditional UWB multi-target positioning.

To address the issues of significant sample imbalance among different quality levels and low grading efficiency in retinal image quality grading tasks, this paper proposes a multi-frequency Transformer-guided graph-based feature aggregation method for retinal image quality grading. First, contrast-limited adaptive histogram equalization (CLAHE) is applied to enhance key details in the images. Then, a ResNet50 network is employed for multi-level feature extraction. Next, a frequency-channel transformer module is designed, which incorporates frequency-domain information to assist in global feature modeling, thereby optimizing the balance between international and local features. Subsequently, a graph cross-feature aggregation module is introduced, leveraging a cross-scale cross-attention mechanism to guide image aggregation, aligning multi-source features, and enhancing the model’s sensitivity to multi-level features. Finally, a weighted loss function increases the model’s attention to minority-class samples. Experiments conducted on the Eye-Quality and RIQA-RFMiD datasets achieved accuracy rates of 88.71% and 84.95%, with precision rates of 87.78% and 74.22%, respectively. The experimental results demonstrate that the proposed algorithm holds significant application value in retinal image quality assessment.

The deformation distribution of engineering structures can evaluate the health status of the bridge structures, but the traditional "point based" sensing technology has significant shortcomings in obtaining the overall structural deformation distribution. This article proposes a curvature identification deflection curve method suitable for monitoring the deformation distribution of simply supported steel box girder structure based on quasi distributed fiber Bragg grating strain measurement technology. Theoretical exploration of the principle of strain curvature deflection curve inversion for fiber Bragg gratings. Numerical simulations of four point bending tests on simply supported box girders show that the method based on curvature identification of deflection has high accuracy, and the relative error at the mid span is only 2.14%. The fiber optic grating strain sensor encapsulated in a stainless steel tube was arranged on the surface of a simply supported steel box girder with a span of 2 meters, and a good linear correlation was obtained between the structural deformation distribution and strain distribution. The maximum relative error between the inverted deflection and the dial gauge at the mid span was less than 3.80%, providing a reference scheme for the long-term deformation monitoring of box girders with distributed distribution.

Hydrogen, as a clean energy source, plays a crucial role in fuel cells and industrial applications. However, its flammability makes the development of real-time, sensitive, and safe hydrogen detection technologies a key factor for expanding its applications. Traditional palladium (Pd)-based thin film sensors, due to their dense structure, have a relatively slow hydrogen diffusion rate and a long response time. To address this issue, this paper proposes the design of a hydrogen optical sensor based on a palladium nanoparticle superlattice thin film, which can be prepared by a self-assembly method to reduce costs. Compared with the traditional dense thin film system, the specific surface area of this palladium nanoparticle thin film can be as high as 1.2×107 cm?1, which is conducive to the rapid diffusion of hydrogen and is expected to improve the sensing response rate. In addition, the calculation results show that when hydrogen diffuses into the interior of palladium to form PdHx, as the atomic ratio (x) of H/Pd increases from 0 to 1, the change in the transmittance intensity of the palladium nanoparticle superlattice thin film reaches 38%, which is significantly greater than that of the traditional dense thin film system, indicating higher sensitivity. This study provides a theoretical basis for the development of real-time and highly sensitive hydrogen sensors.

A simulation system is an efficient tool in laser remote sensing research, effectively advancing system construction and algorithm validation, thereby enhancing research efficiency. To meet the demands of direct scattering spectrum LiDAR for marine remote sensing applications, a comprehensive marine multi-environmental element measurement simulation system was established based on Qt. This system encompasses the laser emission system, seawater channel system, optical reception and spectral measurement system, and signal processing system, achieving full-process simulation. Several simulations were conducted using this system, including system integrity verification and reliability analysis experiments. The simulation results were compared with actual measurement data, showing high consistency and verifying the reliability of the simulation system. The system provides valuable guidance for LiDAR marine remote sensing experiments.

A femtosecond laser with a wavelength of 515 nm is used to perform polishing operations on the surface of titanium metal to analyze the formation mechanism of surface morphology in the processed area under different laser parameters. By changing the laser focus position to suppress the formation of a hump-like structure, the processing quality is significantly improved and the optimal laser parameter window for polished titanium metal is obtained. It is found that when the laser focus is on the surface of the sample, the size of the bump-like structure increases with the increment of the number of processed layers and gradually spreads from the scratch of the sample to the whole processed area. At an energy density of 25.88 J/cm2, the positive defocus amount gradually increases in the range from 0 to 2.5 mm, and then the hump-like structure gradually decreases until it disappears completely, while the polishing depth increases from 38.94 μm to 108.02 μm. The surface roughness is reduced from the original 1.496 μm to 0.314 μm when the defocus amount is 2.5 mm.

3D multi-object tracking technology can help unmanned vehicles accurately perceive their surroundings, identify objects such as pedestrians and vehicles, and predict their movement trajectories, which is important for improving the safety of unmanned systems and reducing traffic accidents. In order to improve the performance of 3D multi-object tracking, a detection-based 3D multi-object tracking framework is proposed, which generates a 3D bounding box with identity, position, and shape in real time from the point cloud provided in each successive frame. Specifically and firstly, to address the shortcomings of traditional IoU, an efficient adaptive conversion intersection and merger ratio (AC-IoU) method is proposed to optimize the data metric between prediction and actual detection. Secondly, in order to increase the number of successful matches and enhance the accuracy of the trajectory, a cascade matching method combining spatial correlation and geometric features is proposed, where the first stage is responsible for matching objects with high confidence, and the second stage focuses on handling objects that are difficult to correlate due to occlusion, low confidence, or other complex situations. Experimental results on the KITTI dataset show that the tracking accuracy is improved by 0.79%, the tracking precision is improved by 3.67%, and the frame rate has reached 115 f/s. These results prove the efficiency and reliability of this paper's method in the task of tracking objects such as pedestrians and cars.

This study addresses the limitations of existing endoscopic optical coherence tomography (OCT) probes in field of view (FOV), resolution, and dimensions constraints by proposing a novel design for a large-FOV forward-scanning endoscopic probe. The probe employs an optical system comprising six spherical lenses with an outer diameter of 1.4 mm, integrated with piezoelectric ceramic-driven fiber scanning technology. Operating in the 1000-1100 nm wavelength range, it achieves a scanning FOV of 6.0 mm, a lateral resolution of 24 μm, and an axial resolution of 20 μm, while maintaining a compact mechanical outer diameter of merely 1.6 mm. A swept-source OCT system was developed to validate the probe’s performance through imaging experiments on ex vivo porcine eyes. Results demonstrate the probe’s capability to clearly distinguish iris, sclera, and retinal structures, as well as visualize retinal detachment phenomena. Comparative analyses with graded-index lens probes reveal superior FOV expansion and reduced imaging artifacts. The study confirms that this probe effectively integrates miniaturization, high-resolution imaging, and large-FOV advantages, providing a new solution for ophthalmic minimally invasive surgical navigation. Future research could focus on further reducing the outer diameter and integrating two-dimensional scanning functionality to enhance its clinical potential.

In the current fine-grained classification task of ships, approaches that rely solely on single image data can only classify by extracting the image features of the target. However, they struggle to capture the complex relationships between the ship's main body and its components, thereby limiting recognition accuracy and results in poor generalization. A data- and knowledge-driven fine-grained classification method, termed DKSCN, is proposed for ships. The object detection network is utilized to detect the ship's main body and its key parts. By designing a graph convolutional network and integrating expert knowledge, a semantic knowledge graph is established to capture the relationships between the ship's main body and its key components. During classification, domain knowledge is incorporated to guide the data-driven process. Comparative experimental results on a self-constructed dataset demonstrate that this method not only addresses the limitations of single data-driven models but also improves classification accuracy.

With the advancement of phased fiber laser array technology toward more units, pointing errors between array beams have become a critical factor affecting beam combining performance. To address the challenges in fabricating light-intensity gradient trap modulation plates for beam pointing control based on light-intensity gradient traps, this paper proposes and employs a lithographic mask process to fabricate a light-intensity gradient trap modulation plate with annular stripe patterns. The measured light intensity transmittance aligns with the design value, and the root mean square error (RMSE) of the normalized transmittance is 0.0166. Through simulations and experiments, rapid correction of beam pointing errors in an extended area was achieved using the stochastic parallel gradient descent algorithm and an adaptive fiber-optics collimator. With an initial average beam pointing error of approximately 702 μrad, the mean residual pointing error after correction was reduced to less than 10 μrad. This research provides important references for laser pointing control and beam combining on extended targets.

Synchrotron radiation-based full-field nano-computed tomography (CT) is capable of non-destructive detection of the internal 3D structure of samples with nanoscale spatial resolution, and has a wide range of applications. Conventional nano-CT requires the acquisition of numerous projection images to ensure the accuracy and high resolution of 3D reconstruction, which is not only time-consuming but also may cause radiation damage to the sample. In this study, a novel denoising network model, SwinCBD (Swin Transformer-based convolutional blind denoising), is proposed to address the challenges of nano-CT technology. The SwinCBD model is based on Swin Transformer and convolutional neural network to establish structural relation mapping between noisy images and clean images through deep learning to achieve high-quality reconstruction of low-dose nano-CT with low exposure and fewer projections. The experimental results show that the low-dose nano-CT image denoising based on the SwinCBD model improves the signal-to-noise ratio of the low-dose CT slice images (by 49.26%), and drastically reduces the nano-CT projections acquisition time under the premise of ensuring the image quality. The model will be important for improving the nano-CT temporal resolution and reducing the radiation damage of samples.

UAV aerial images have the characteristics of complex background, small and dense targets. Aiming at the problems of low precision and a large number of model parameters in UAV aerial image detection, an efficient multi-scale feature transfer small target detection algorithm based on hypergraph computation is proposed. Firstly, a multi-scale feature pyramid network is designed as a neck network to effectively reduce the problem of information loss caused by lengthy transmission paths by fusing multi-layer features in the middle layer and transmitting them directly to adjacent layers. In addition, the feature fusion process uses hypergraphs to model higher-order features, improving the nonlinear representation ability of the model. Secondly, a lightweight dynamic task-guided detection head is designed to effectively solve the problem of inaccurate detection targets caused by inconsistent classification and positioning task space in the traditional decoupling head with a small number of parameters through sharing mechanism. Finally, the pruning lightweight model based on layer adaptive amplitude is used to further reduce the model volume. The experimental results show that this algorithm has better performance than other architectures on VisDrone2019 dataset, with the accuracy mAP0.5 and parameter number reaching 42.4% and 4.8 M, respectively. Compared with the benchmark YOLOv8, the parameter number is reduced by 54.7%. The model achieves a good balance between detection performance and resource consumption.

Optical bubble, characterized by a tightly focused three-dimensional dark-field intensity distribution, exhibits significant application value in fields such as optical manipulation and laser processing. In previously reported results, an optical bubble is typically generated through multi-beam interference and superposition, which involves complex optical setups and is not conducive to system integration and practical applications, and has low energy utilization efficiency. In this study, we utilize single-beam vector field modulation technology to generate a tightly focused optical bubble with high intensity uniformity. Furthermore, we achieve the detection of this hollow bubble through polarization conversion of the probe light. By adjusting the energy ratio between azimuthally polarized incident beam and radially polarized incident beam modulated by a 0/π binary phase, we experimentally realize an optical bubble with an edge-to-center dark spot intensity ratio exceeding 10:1 and edge intensity uniformity approaching 90%. This work provides a feasible technical approach for applications in dual-beam super-resolution laser processing, optical data storage, and particle manipulation.

In order to improve the detection capability of small target defects in steel surface inspection, an improved YOLOv8-SOE model is proposed. The model processes the P2 layer features by designing the FSCConv module. By compressing the P2 layer features and deeply fusing them with the P3 layer features, the model's sensitivity to small target features is effectively enhanced, while avoiding the computational burden caused by the introduction of additional detection layers. In order to further optimize the multi-scale feature fusion capability, cross stage partial omni-kernel (CSP-OK) module is used to optimize the multi-scale feature fusion, which improves the integration efficiency of features of different scales. The SIoU loss function is introduced to optimize the bounding box regression, which further improves the positioning accuracy. Experimental results show that the mAP of the YOLOv8-SOE model on the NEU-DET dataset achieves 80.7%, which is 5.4% higher than the baseline model, and has good generalization ability on the VOC2012 dataset. While improving the accuracy of small target detection, the model maintains a high computational efficiency and has good application prospects.

Electrostatically driven membrane deformable mirrors correct wavefront aberrations through electrostatic forces, whose correction capability dependent on driving load accuracy. Due to the charge aggregation at the electrode edges, non-uniform deformation occurs due to the nonlinear change of the regional load, which causes inaccurate or incorrect assessment of the calibrated wavefront aberration of electrostatically driven membrane deformable mirrors. Based on this, this work carries out a study on electrostatically driven membrane deformable mirror electrode edge effect and its influence on correction capability assessment, establishes a theoretical model of electrode edge effect, and quantitatively analyzes the influence of electrode edge deformation response and correction capability assessment accuracy based on the model, and the results show that before and after the consideration of the edge effect, the error of correction capability assessment is improved from 25.49% to 6.83% or even lower, and applies to different electrode spacing parameters, which verifies the correctness of the theoretical model proposed in this paper.

This paper presents a hybrid multi-order diffractive lens that supports dual-band computational imaging in both visible and mid-wave infrared (MWIR) bands. By applying diffraction structures of varying depths on both sides of the same substrate and optimizing these structural parameters using the end-to-end optimization framework, we successfully developed a diffractive element capable of efficient focusing in the visible light band (640~800 nm) and the MWIR band (3700~4700 nm). Coupled with a specially designed image reconstruction network, this approach realizes a monolithic dual-band computational imaging system with simplicity, lightweight construction, and low cost. Experimental results show that the prototype with a diameter of 40 mm achieves static modulation transfer functions of 50.0% in the visible light band and 4.4% in the MWIR band. Under room temperature conditions, the noise equivalent temperature difference in the infrared band does not exceed 80 mK, confirming the effectiveness and practicality of the proposed design.

To address the challenge of prolonged acquisition times in magnetic resonance imaging (MRI), data-driven algorithms and the model integration have emerged as crucial approaches for achieving high-quality MRI reconstruction. However, existing methods predominantly focus on visual feature extraction while neglecting deep semantic information critical for robust reconstruction. To bridge this gap, this study proposes a model-driven architecture that synergistically combines hierarchical semantic networks with physical model networks, aiming to enhance reconstruction performance while maintaining computational efficiency. The architecture comprises four core modules: a context extraction module to capture rich contextual features and mitigate background interference; a multi-scale aggregation module integrating multi-scale information to preserve coarse-to-fine anatomical details; a semantic graph reasoning module to model semantic relationships for improved tissue differentiation and artifact suppression; a dual-scale attention module to enhance critical feature representation across different detail levels. This hierarchical and semantic-aware design effectively reduces aliasing artifacts and significantly improves image fidelity. Experimental results demonstrate that the proposed method outperforms state-of-the-art approaches in both quantitative metrics and visual quality across diverse datasets with varying sampling rates. For instance, in 4× radial acceleration experiments on the IXI dataset, our approach achieved a peak signal-to-noise ratio (PSNR) of 48.15 dB, surpassing the latest comparison algorithms by approximately 1.00 dB on average, while enabling higher acceleration rates and maintaining reliable reconstruction outcomes.

The flat mirror with the characteristics of large aspect ratio and high lightweight rate is one difficulty in the opto-mechanical design of a large off-axis three-mirror anastigmat cameras. For a certain flat mirror with a clear aperture of 1220 mm×198 mm, the assembly structure combining a semi-closed mirror blank made of silicon carbide with the three-point back support scheme was proposed, resulting in a total design weight of 30.5 kg. The supporting effect of the mirror was improved through the optimization of support positions. Both the size and position of hinges in the flexure were adjusted, taking into account gravitational deformation, thermal stability, and dynamic characteristics of the assembly. Simulation reveals that, under the condition of gravity during the test, the root mean square (RMS) of the surface accuracy change of the flat mirror is 1.812 nm, together with the tilt of 3.639" for the mirror blank. The measured fundamental frequency of the assembly is 132.5 Hz. After polishing, the tested RMS values of surface accuracy are 0.0203λ, 0.0197λ, and 0.0204λ (λ=632.8 nm), corresponding to the left, middle, and right sub-zones of the flat mirror respectively. The surface accuracy can remain basically unchanged after environmental tests, which meets the requirements of high-performance space cameras.

To address the challenges of missed detection and false detection caused by complex backgrounds, varying illumination, target occlusion, and scale diversity in UAV images, this paper proposes a multi-level refined object detection algorithm for UAV imagery. First, a CSP-SMSFF (cross-stage partial selective multi-scale feature fusion) module is designed by integrating multi-scale feature extraction and feature fusion enhancement strategies. This module employs incremental convolutional kernels and channel-wise fusion to precisely capture multi-scale target features. Second, an AFGCAttention (adaptive fine-grained channel attention) mechanism is introduced, which optimizes channel feature representations through a dynamic fine-tuning mechanism. This enhances the algorithm’s sensitivity to critical multi-scale sample features, improves discriminative capability, preserves fine-grained mapping information, and suppresses background noise to mitigate missed detection. Third, a SGCE-Head (shared group convolution efficient head) detection head is developed, leveraging EMSPConv (efficient multi-scale convolution) to achieve precise capture of global salient features and local details in spatial-channel dimensions, thereby enhancing localization and recognition of multi-scale features and reducing false positives. Finally, the Inner-Powerful-IoUv2 loss function is proposed, which balances localization weights for samples of varying quality through dynamic gradient weighting and hierarchical IoU optimization, thereby strengthening the model’s capability to detect ambiguous targets. Experimental results on the VisDrone2019 and VisDrone2021 datasets benchmark demonstrate that the proposed method achieves 47.5% and 45.3% in mAP@0.5 under two evaluation settings, surpassing baseline models by 5.7% and 4.7%, respectively, and outperforming existing comparative algorithms.

To achieve high-precision measurement of the all-Stokes vector, a mutual differential frequency modulation approach for dual photo-elastic modulators (photoelectric modulator, PEM) in all-Stokes vector measurement was proposed. The incident light was differently frequency modulated by two PEMs of different frequencies. The measured polarization vector and the phase delay amplitude of the PEM were simultaneously modulated at different differential frequency components. The phase delay amplitude of the PEM was obtained in real time by dividing the odd differential frequency components, and the Stokes vector of the measured light was accurately obtained by combining different differential frequency components. This method reduced the error introduced by the fluctuation of the phase delay amplitude in the PEM measurement system. Theoretical and experimental analyses were conducted, and the results showed that the variance of the measured Stokes vector was 10?5. This method will provide support for high-precision polarization measurement.

Aiming at the problem that the single-stage YOLACT algorithm based on bounding box detection lacks the location and extraction of the region of interest, and the issue that two bounding boxes overlap and are difficult to distinguish, this paper proposes an anchor-free instance segmentation method based on the improved YOLACTR algorithm. The mask generation is decoupled into feature learning and convolution kernel learning, and the feature aggregation network is used to generate mask features. By adding position information to the feature map, multi-layer transformer and two-way attention are used to obtain dynamic convolution kernels. The experimental results show that this method achieves a mask accuracy (AP) of 35.2% on the MS COCO public dataset. Compared with the YOLACT algorithm, this method improves the mask accuracy by 25.7%, the small target detection accuracy by 37.1%, the medium target detection accuracy by 25.8%, and the large target detection accuracy by 21.9%. Compared with YOLACT, Mask R-CNN, SOLO, and other methods, our algorithm shows significant advantages in segmentation accuracy and edge detail preservation, especially excelling in overlapping object segmentation and small target detection, effectively solving the problem of incorrect segmentation in instance boundary overlap regions that traditional methods face.

To address the challenge of balancing the computational accuracy and efficiency in adaptive finite element meshing, this study proposes a GTF-Net model based on the attention fusion mechanism. The model combines the graph attention network with the Transformer architecture, dynamically couples local geometric features with the global physical field through a multi-head cross-attention module, and enhances the representation of singular fields and complex boundaries. The verification of two case studies of waveguide transmission and Bessel equation shows that compared with the traditional Scikit-FEM (skFem) method, GTF-Net improves computational efficiency while reducing the standard deviation of gradient error by 85.9% and 23.8%, respectively. The results show that the model significantly improves the fit between mesh distribution and physical field changes through nonlinear feature mapping, providing a novel deep learning solution for adaptive mesh optimization in engineering calculations.

In response to the current limitations of neck pulse monitoring devices, such as being inconvenient to carry and having complex signal processing, a fiber Bragg grating (FBG) based neck pulse monitoring device was designed. The device monitored two volunteers in three states (resting, exercise, and vigorous exercise) while sitting and lying down for 10 s each. Fourier transform was applied to process the data, and the frequency error between the neck pulse device, the wristband, and the pulse oximeter was found to be less than 10%. Pearson correlation analysis was conducted on the periods of different states, with the correlation coefficient exceeding 0.9. Random forest was used for predictive analysis, and the results showed good prediction performance. The analysis indicates that the neck pulse monitoring device is capable of effectively monitoring the pulse in the neck region of the human body.

With the wide application of point clouds in virtual reality, computer vision, robotics and other fields, the assessment of distortions resulted from point cloud acquisition and processing is becoming an important research topic. Considering that the three-dimensional information of point clouds is sensitive to geometric distortion and the two-dimensional projection of point clouds contains rich texture and semantic information, a no-reference point cloud quality assessment method based on the fusion of three-dimensional and two-dimensional features is proposed to effectively combine the three-dimensional and two-dimensional feature information of point cloud and improve the accuracy of point cloud quality assessment. For 3D feature extraction, the farthest point sampling is firstly implemented on the point cloud, and then the non-overlapping point cloud sub-models centered on the selected points are generated, to cover the whole point cloud model as much as possible and use a multi-scale 3D feature extraction network to extract the features of voxels and points. For 2D feature extraction, the point cloud is first projected with orthogonal hexahedron projection, and then the texture and semantic information are extracted by a multi-scale 2D feature extraction network. Finally, considering the process of segmentation and interweaving fusion that occurs when the human visual system processes different types of information, a symmetric cross-modal attention module is designed to integrate 3D and 2D features. The experimental results on five public point cloud quality assessment datasets show that the Pearson’s linear correlation coefficient (PLCC) of the proposed method reaches 0.9203, 0.9463, 0.9125, 0.9164 and 0.9209 respectively, indicating that the proposed method has advanced performance compared with the existing representative point cloud quality assessment methods.

This paper proposed a multi-task attention mechanism-based no-reference quality assessment algorithm for screen content images (MTA-SCI). The MTA-SCI first used a self-attention mechanism to extract global features from screen content images, enhancing the representation of overall image information. It then applied an integrated local attention mechanism to extract local features, allowing the focus to be on attention-grabbing details within the image. Finally, a dual-channel feature mapping module predicted the quality score of the screen content image. On the SCID and SIQAD datasets, MTA-SCI achieves Spearman's rank-order correlation coefficients (SROCC) of 0.9602 and 0.9233, and Pearson linear correlation coefficients (PLCC) of 0.9609 and 0.9294, respectively. The experimental results show that the MTA-SCI achieves high accuracy in predicting screen content image quality.

With the rapid development of convolutional neural networks (CNNs) and Transformer models, significant progress has been made in remote sensing image super-resolution (RSSR) reconstruction tasks. However, existing methods have limitations in effectively handling multi-scale object features and fail to fully explore the implicit correlations between channel and spatial dimensions, thus restricting further improvements in reconstruction performance. To address these issues, this paper proposes an adaptive dual-domain attention network (ADAN). The network integrates self-attention information from both channel and spatial domains to enhance feature extraction capabilities. A multi-scale feed-forward network (MSFFN) is designed to capture rich multi-scale features. At the same time, an innovative gated convolutional module is introduced to further enhance the representation of local features. The network adopts a U-shaped backbone structure, enabling efficient multi-level feature fusion. Experimental results on multiple publicly available remote sensing datasets show that the proposed ADAN method significantly outperforms state-of-the-art approaches in terms of quantitative metrics (e.g., PSNR and SSIM) and visual quality. These results validate the effectiveness and superiority of ADAN, providing novel insights and technical approaches for remote sensing image super-resolution reconstruction.

To address the issue of complex backgrounds in dim scenes, which cause object edge blurring and obscure small objects, leading to misdetection and omission, an improved YOLOv8-GAIS algorithm is proposed. The FAMFF (four-head adaptive multi-dimensional feature fusion) strategy is designed to achieve spatial filtering of conflicting information. A small object detection head is incorporated to address the issue of large object scale variation in aerial views. The SEAM (spatially enhanced attention mechanism) is introduced to enhance the network's attention and capture ability for occluded parts in low illumination situations. The InnerSIoU loss function is adopted to emphasize the core regions, thereby improving the detection performance of occluded objects. Field scenes are collected to expand the VisDrone2021 dataset, and the Gamma and SAHI (slicing aided hyper inference) algorithms are applied for preprocessing. This helps balance the distribution of different object types in low-illumination scenarios, optimizing the model's generalization ability and detection accuracy. Comparative experiments show that the improved model reduces the number of parameters by 1.53 MB, and increases mAP50 by 6.9%, mAP50-95 by 5.6%, and model computation by 7.2 GFLOPs compared to the baseline model. In addition, field experiments were conducted in Dagu South Road, Jinnan District, Tianjin City, China, to determine the optimal altitude for image acquisition by UAVs. The results show that, at a flight altitude of 60 m, the model achieves the detection accuracy of 77.8% mAP50.

Disturbance suppression, especially high-frequency disturbance suppression beyond the closed-loop bandwidth, is the core of realizing high-precision stability control for tip-tilt correction systems. Repetitive control has good performance of periodic trajectory tracking and disturbance suppression, which is applied to the stability control of high-precision systems. The high-frequency disturbance suppression problem of the tip-tilt correction system is analyzed in this paper, and the performance of high-frequency disturbance suppression based on repetitive control is summarized. To solve the problems of natural frequency drift and waterbed amplification in traditional repetitive controllers, a comb-like repetitive controller based on Youla parameterization is designed to suppress high-frequency disturbances beyond the closed-loop bandwidth. In order to solve the problem that the integer-order repetitive control is only effective for specific frequency points, especially in most high frequency regions, the controller will fail due to disturbance fluctuations and uncertainty, an all-pass frit-order delay filter is optimized to suppress the high frequency disturbance at any frequency point up to Nyquist frequency in the tip-tilt correction system. Finally, a parallel repetitive control scheme is designed to suppress the vibration of aperiodic structures which is difficult to suppress, and its robust stability and effectiveness are discussed.

Aiming at the problems of temperature sensitivity, setting at room temperature, defocusing at low temperature, and low accuracy of camera focal surface prefabrication at room temperature, a defocusing test technology of low temperature infrared optical system based on opto-machine co-simulation, interferometry, and linear displacement measurement was proposed. The main factors causing low temperature defocusing are analyzed, and the defocusing data is calculated by optical machine simulation. The low temperature interference test optical path of the optical system is established by using the sensitive characteristic of power to focus position, and the low temperature defocusing of the infrared optical system is tested by combining high precision displacement measurement. This technology is used to analyze and test the defocus of a light and small infrared camera with an 181 mm aperture from normal temperature to low temperature. The deviation between the test results and the simulation calculation is less than half of the system focal depth. On this basis, the camera is preset at a normal temperature focal plane, and the experiment shows that the preset focal plane is accurate, which proves the feasibility and accuracy of the low temperature defocus measurement method. The test method can be used for presetting and focusing light and a small optical camera at normal temperature which is sensitive to temperature.

Aiming at the problem of leakage and misdetection caused by the high percentage of sample overlapping and occlusion, the difficulty of key feature extraction, and the large background noise in X-ray security images, an adaptive panoramic focusing X-ray image contraband detection algorithm is proposed. Firstly, the foreground feature awareness module is designed to accurately distinguish contraband and background noise by enhancing the edge structure and texture details of the foreground target to improve the accuracy and completeness of feature representation. Then, the multi-path two-dimensional information integration module is constructed by combining the multi-branch structure and dual cross attention mechanism to optimize the feature interaction and fusion in the channel and spatial dimensions, to strengthen the extraction capability of key features, and to effectively suppress the background interference. Finally, a panoramic dynamic focus detection head is constructed, which dynamically adjusts the receptive field through frequency adaptive dilated convolutions to accommodate the feature frequency distribution of small-sized contraband targets, thereby enhancing the model's ability to recognize small targets. Trained and tested on the public datasets SIXray and OPIXray, the mAP@0.5 reaches 93.3% and 92.5%, respectively, outperforming the other compared algorithms. The experimental results show that the proposed model significantly improves the leakage and false detection of contraband in X-ray images with high accuracy and robustness.

To address the challenges of uneven inter-class distribution and difficulty in lesion area recognition in retinal fundus image datasets, this paper proposes a fusion dual-attention retinal disease grading algorithm with PVTv2 and DenseNet121. First, retinal images are preliminarily processed through a dual-branch network of PVTv2 and DenseNet121 to extract global and local information. Next, spatial-channel synergistic attention modules and multi-frequency multi-scale attention modules are applied to PVTv2 and DenseNet121, respectively. These modules refine local feature details, highlight subtle lesion features, and enhance the model's sensitivity to complex micro-lesions and its spatial perception of lesions areas. Subsequently, a neuron cross-fusion module is designed to establish long-range dependencies between the macroscopic layout and microscopic texture information of lesion areas, thereby improving the accuracy of retinal disease grading. Finally, a hybrid loss function is employed to mitigate the imbalance in model attention across grades caused by uneven sample distribution. Experimental validation on the IDRID and APTOS 2019 datasets yields quadratic weighted kappa scores of 90.68% and 90.35%, respectively. The accuracy on the IDRID dataset and the area under the ROC curve on the APTOS 2019 dataset reached 80.58% and 93.22%, respectively. The experimental results demonstrate that the proposed algorithm holds significant potential for application in retinal disease grading.

To address the poor visibility of submerged bleeding points during endoscopic surgery, which can extend hemostasis procedures, a method combining 4-LED illumination with a saturation-to-hue mapping algorithm is proposed for enhancing bleeding points visibility. With imaging specificity for tissue and blood of varying concentrations, four narrow-band LEDs synchronous lighting are used for imaging, replacing the conventional xenon light source. In the xyY color space, the saturation information of the image is stretched and then mapped to the hue, effectively increasing the color differences between submerged bleeding points and the surrounding areas. Vitro experimental results using rigid endoscopes on animal materials demonstrate that the enhanced images achieved significantly higher color difference between bleeding points and their surroundings than conventional white-light images (41.31>11.78). Saturation-to-hue mapping imaging (SHMI) effectively improves the visibility of submerged bleeding points and reduces the risk associated with endoscopic surgeries.

Increasing the active vibration isolation capability between the optical payload and the motion platform has always been a challenge for optoelectronic tracking systems. Therefore, a dual observer method is proposed to achieve wide-band disturbance rejection for an inertially stabilized platform. The dual observer method consists of two aspects. Firstly, a classical error observer has a strong low-frequency suppression ability through the design of a low-pass filter. Secondly, a saturated acceleration disturbance observer improves its disturbance suppression characteristics and completes the rejection of medium and high-frequency disturbances by adjusting the saturation threshold and filter bandwidth according to its stability conditions. The dual observer combines both advantages, and the interaction between the two observers is analyzed for better parameterization. Closed-loop verification of the proposal is carried out using the inertial stabilization device. The experimental results show that the dual observer can improve the closed-loop performance under both single-frequency and mixed-frequency disturbances.

In response to the problems of traditional defect detection algorithms, such as poor accuracy and feature loss in practical applications due to the inconspicuous characteristics of welding defects and complex background information, this paper proposes a welding surface defect detection algorithm based on the improved YOLOv8 (GD-YOLO). The model first introduces the fusion of feature extraction modules and convolutional modules to enhance its information extraction capabilities. Then, a slim-neck structure is embedded in the neck network, and the upsampling operator CAFARE is referenced in the feature fusion stage to assist in enhancing the model's performance. Subsequently, the attention mechanism module is improved to optimize the overall performance without significantly increasing the computational burden. Finally, the loss function is changed to Inner-SIOU to address the problem of mismatched bounding boxes. Experimental results show that the mAP0.5 detection metric of the model in this paper is 7.8% higher than that of the baseline model, and the number of parameters and the amount of computation are reduced by 0.2 M and 0.7 G, respectively.

In response to the on-orbit reconfiguration challenges faced by future large aperture segmented optical systems, a lightweight design method with a wide range of curvature adjustability is proposed. This study initially analyzes the relationship between the characteristics of piezoelectric ceramics and the constitutive equations of thermal strain, deducing that piezoelectric strain can be precisely equivalent to thermal strain. Based on the flexural curve equation, the deformation of piezoelectric ceramics is calculated, enabling the parameterized modeling of an ultra-low expansion (ULE) glass segmented mirror with an edge distance of 510 mm and a curvature radius of 9000 mm. Simulation results indicate that 54 interlaced actuators can achieve a curvature radius reconfiguration of 240.07 mm with a control voltage range of ±20 V, exhibiting a highly linear relationship. Experimental results further demonstrate that when the control voltage is varied between -25 V and 20 V, the change in the curvature radius of the segmented mirror reaches 233.44 mm, with the positive unit voltage corresponding to a greater change in curvature radius than the negative. The proposed method for a wide range of curvature adjustable segmented mirrors provides new insights for the engineering application of large aperture segmented optics in on-orbit reconfiguration.

The accurate identification of the hip joint keypoint is vital for diagnosing developmental dysplasia of the hip. However, in pediatric hip X-ray images, bone regions around key points often exhibit low contrast and blurred edges, resulting in unclear edge features. Furthermore, down-sampling operations during feature extraction further weaken edge information. Key structures surrounding the keypoint are highly susceptible to background interference. Such factors hinder the precise localization of key points. An edge feature and detail-aware integrated YOLOv8s algorithm was proposed for hip joint key point detection. The algorithm designs an edge feature enhancement module to capture spatial information around key points and strengthen edge features. A detail-aware network was designed to integrate and refine multi-level features, enhancing image perception of fine structures. Experiments used a hip X-ray dataset from the Department of Radiology, Children's Hospital of Chongqing Medical University. Results showed reductions in average keypoint localization and angular errors to 4.2090 pixel and 1.4872°, respectively. These reductions, which are 6.8% and 9.9% compared to those of YOLOv8s, highlight significant improvements in detection accuracy. The algorithm enhances keypoint detection precision and provides valuable support for clinical diagnosis.

To address challenges such as regional mis-segmentation and insufficient target localization accuracy in colorectal polyp segmentation, this paper proposes a colorectal polyp segmentation algorithm that integrates adaptive feature selection based on a Transformer. Firstly, the Transformer encoder is employed to extract multi-level feature representations, capturing multi-scale information from fine-grained to high-level semantics. Secondly, a dual-focus attention module is designed to enhance feature representation and recognition capabilities by integrating multi-scale information, spatial attention, and local detail features, significantly improving the localization accuracy of lesion areas. Thirdly, a hierarchical feature fusion module is introduced, which adopts a hierarchical aggregation strategy to strengthen the fusion of local and global features, enhancing the capture of complex regional features and effectively reducing mis-segmentation. Finally, a dynamic feature selection module is incorporated with adaptive selection and weighting mechanisms to optimize multi-resolution feature representation, eliminate redundant information, and focus on key areas. Experiments conducted on the Kvasir, CVC-ClinicDB, CVC-ColonDB, and ETIS datasets achieved Dice coefficients of 0.926, 0.941, 0.814, and 0.797, respectively. The experimental results demonstrate that the proposed algorithm exhibits superior performance and application value in the task of colorectal polyp segmentation.

The low detection rate of tiny defects on the surface of metal pipe fittings is a key issue confronting industrial component inspection. In aiming at this problem, an improved YOLOv9-MM model was constructed to improve the accuracy of small target detection. A real-time image acquisition system for precision metal pipe fittings was designed. By using an annular light source combined with a telecentric lens, the surface of pipe fittings can be snapped by the CCD camera and covered at all angles to eliminate the problem of missing areas. The feature map extracted methods of shallow network were introduced, and the upper sampling module of Dysample was combined to realize the dynamic fusion of depth features. By improving the loss function, the precision of small target detection is greatly improved. The results show that the proposed method has an average detection accuracy of 70.2% and a detection speed of 90 f/s. The proposed method shows some feasibility in the actual application.

To reduce the impact of random errors on hand-eye calibration in the visual system of a blade repair robot, an optimization method based on outlier detection is proposed. Firstly, a linear equation for the hand-eye matrix is established. The initial hand-eye matrix is solved using singular value decomposition (SVD). Secondly, the initial value is used to perform an inversion operation on the samples. Outlier samples are detected and removed based on Z-scores, leading to a more accurate hand-eye matrix. Finally, the obtained hand-eye matrix is used as the initial value for optimization. The rotation is represented by unit quaternions, and the Levenberg-Marquardt algorithm is applied to further optimize the initial value, yielding the final hand-eye matrix. Hand-eye calibration experiments were conducted on the blade repair robot equipped with a stereo depth camera. The real coordinates of the target points were obtained using a TCP calibration tool. The predicted coordinates from the hand-eye matrix, obtained by the proposed method, have an average Euclidean distance of 0.858 mm from the true coordinates, with a variance stabilizing below 0.1. Compared to other methods, the proposed approach effectively reduces the impact of random errors and demonstrates good stability and accuracy.

The Risley grating tracking system is mainly composed of two rotating polarization gratings. The light source is diffracted by the polarization grating to achieve beam pointing in the conical range, and then the target is captured and tracked. As an important index of the Risley grating tracking system, pointing accuracy is not only affected by servo and optical systems but also by system errors such as antenna installation accuracy and shafting assembly error of double grating turntable in the Risley grating tracking system. Therefore, this paper mainly analyzes the systematic error sources in the Risley grating tracking system and the pointing errors caused by them. First, a mathematical model of systematic error is established and verified by ZEMAX. Then, MATLAB is used to analyze the influence of each systematic error source on the pointing error of the Risley grating tracking system. Finally, according to the analysis results and index requirements, the error source of a double grating tracking system is assigned to guide the design and installation of the double grating turntable. The actual maximum pointing error of the double grating turntable δe=7.2" is obtained after several experimental tests, which satisfies the design index of pointing error of the double grating turntable 10".

Aiming at the poor segmentation effect caused by the large scale difference of objects, uneven spatial distribution of samples, fuzzy boundary of objects and large span of scene area, this paper proposes a high-precision remote sensing building segmentation algorithm enhanced by integrating dynamic features. Firstly, the New_GhostNetV2 network is constructed, and the adaptive context-aware convolution is used to improve the algorithm's ability to capture the features of the sample space. Secondly, multi-level information enhancement modules are designed using ghost convolution combined with skip connections and feature branching strategies to enhance the feature integration. Then CGA (cascaded group attention) is introduced to enhance the adaptability of the model to diverse ground object forms through the calculation of independent attention within the group. Finally, the feature fusion module is constructed by the dynamic depth feature enhancer to further enhance the ability of model capture. The experimental results on the WHU data set show that the improved algorithm is 8.57% higher than the baseline model F1-Score and 12.48% higher than mIoU. Compared with other mainstream semantic segmentation models, the improved DeepLabv3+ has better segmentation accuracy.

In 2024, under the background of the National Natural Science Foundation of China (NSFC) reform of the project application rules and scientific research funding management mechanism during the centralized acceptance period, the discipline of F05 "Optics and Optoelectronics" carried out funding strategy reform and practical work. Firstly, this work focuses on the application and funding situation of general projects, youth science fund projects, regional science fund projects, key projects, outstanding youth science fund projects and outstanding young scholar science fund projects, and systematically analyzes the application scale, funding ratio and change trend. Then, the distribution of project applications and funding support units is comprehensively sorted out. And the current frontier research hotspots and potential growth points of optics and optoelectronics are analyzed in combination with the classification of secondary codes. Finally, the comparative effect of the "responsible, credibility, and contribution" (RCC) evaluation mechanism on the fairness, justice and efficiency of the discipline funding system is revealed. The purpose of this work is to provide a detailed reference framework for experts and scholars who apply for NSFC funding in the next year and stimulate innovation.

In response to the deficiencies of existing steel surface defect detection algorithms in terms of resource consumption, detection accuracy, and efficiency, a lightweight steel defect detection algorithm based on YOLOv8n (FCM-YOLOv8n) is proposed. First, a frequency-aware feature fusion network is utilized to efficiently extract and integrate high-frequency information, reducing computational costs while enhancing detection speed. Second, a lightweight feature interaction module (Cc-C2f) is restructured to effectively preserve spatial and channel dependencies while reducing feature redundancy, thereby lowering model parameters and computational complexity. Finally, a multi-spectrum attention mechanism is applied to mitigate feature information loss in the frequency domain, improving the accuracy of detecting complex defects. Experimental results on the Severstal and NEU-DET steel defect datasets show that, compared to YOLOv8n, the FCM-YOLOv8n algorithm achieves a 2.2% and 1.5% improvement in mAP@0.5, respectively, with a 0.5 M and 1.5 G reduction in parameters and computational complexity. The FPS reaches 143 f/s and 154 f/s, respectively, demonstrating excellent real-time performance. The algorithm achieves an optimal balance between detection accuracy, computational cost, and efficiency, providing robust support for edge device applications. Further validation on the GC10-DET dataset shows a 2.9% improvement in mAP@0.5 compared to the baseline model, fully demonstrating the algorithm's exceptional generalization ability.

In this research, an all-dielectric terahertz metasurface based on bound states in the continuum (BIC) is proposed. Each structural unit of the metasurface consists of two rectangle blocks with square cross-sections and a substrate. The substrate material is quartz, and the rectangle block material of the surface is lossless silicon. The symmetry of the metasurface is broken by tchanging the cross-section area of the rectangle block, and the quasi-BIC is excited. The resonance with extremely narrow linewidth is obtained. The transmission spectra with different asymmetric, structural and material parameters are studied using finite element method (FEM) and control variable method. Meanwhile, the Q factor of the proposed metasurface is calculated, which can reach 1.1006×104 and is higher than the Q factors from related listed references. In addition, this study is aimed at the limitations of the relatively limited research on the equivalent parameters of all-dielectric metasurfaces, the S-parameter extraction method is utilized to calculate and analyze the equivalent parameters of the proposed metasurface and the physical properties of the metasurface is studied from this perspective preliminarily.

Aiming at the issues of low recognition accuracy, high computational cost, and large model size caused by the complex background and small target scale in remote sensing images of military aircraft, a lightweight military aircraft target detection algorithm, namely YOLOv8-MA, integrating reparameterization and detail enhancement is proposed. Firstly, a multi-branch gradient flow feature extraction module is designed through reparameterization to enhance the model's inference speed. Secondly, in combination with efficient RepGFPN, redundant model structures are discarded and the P2 layer is incorporated to construct a multi-scale feature fusion network, mitigating the problem of small target information loss due to excessive downsampling. On this basis, a lightweight detection head is proposed by integrating GN convolution and detail enhancement to reduce the number of model parameters and the amount of computation. Finally, a focus coefficient is introduced into the Shape-IoU to form a new loss function, thereby improving the detection performance of the model. On the public military aircraft dataset MAR20, the mAP50 of this algorithm is as high as 97.9%, and the model size is as low as 2.1 MB. Compared with YOLOv8n, the number of parameters decreases by 74.7%, the amount of computation reduces by 40.7%, and the FPS increases by 14 f/s, demonstrating that it can effectively enhance the detection effect of military aircraft in remote sensing images.

To address the challenges posed by complex interference in retinal microsurgery, this study presents a deep learning-based algorithm for surgical instrument detection. The RET1 dataset was first constructed and meticulously annotated to provide a reliable basis for training and evaluation. Building upon the YOLO framework, this study introduces the SGConv and RGSCSP feature extraction modules, specifically designed to enhance the model's capability to capture fine-grained image details, especially in scenarios involving degraded image quality. Furthermore, to address the issues of slow convergence in IoU loss and inaccuracies in bounding box regression, the DeltaIoU bounding box loss function was proposed to improve both detection precision and training efficiency. Additionally, the integration of dynamic and decoupled heads optimizes feature fusion, further enhancing the detection performance. Experimental results demonstrate that the proposed method achieves 72.4% mAP50-95 on the RET1 dataset, marking a 3.8% improvement over existing algorithms. The method also exhibits robust performance in detecting surgical instruments under various complex surgical scenarios, underscoring its potential to support automatic tracking in surgical microscopes and intelligent surgical navigation systems.

A kind of method of structural optimization design using structural deformation to compensate for optical misalignment is presented, which aims at decreasing the affection of imaging by gravity. First, on the basis of the characteristics of the optical system of a space camera, the structural forms of the main frame of the camera and the flexible support structure of the mirrors are determined. Secondly, taking the relative stiffness shift, relative inclination and face shape error of the secondary mirror, tertiary mirror, folding mirror, focusing mirror and main mirror under gravity as the optimization objectives, and the stability tolerance requirement of the optical system for each mirror as the constraints, the optimal parameters of the truss of main frame, the bearing cylinder thickness and the flexible support structure are obtained. Finally, the finite element analysis is carried out on the mode and Surface deformation of the camera, and sinusoidal sweep test and system wave aberration test are carried out on the camera. The results show that the first-order frequency of the camera is 36.7 Hz, and the average wave aberration of each field of the camera is 0.0655λ and 0.0701λ (λ=632.8 nm) under the two states of loading and flipping 180 degrees, respectively. It can be inferred that the wave aberration of the system can meet the optical design requirements after the gravity effect of the camera is disappeared. The optimized design of the camera support structure can be used as a reference for the structural design of space cameras.

To address the challenges of background complexity and target scale changes in synthetic aperture radar (SAR) images, especially in densely populated small-target scenes prone to false and missed detections, a multi-granularity feature and shape-position similarity metric method for ship detection in SAR images is proposed. First, a multi-granularity feature aggregation structure containing two branches is designed in the feature extraction stage. One branch decomposes the feature map cascade by Haar wavelet transform to expand the global receptive field to extract coarse-grained features. The other branch introduces spatial and channel reconstruction convolution to capture detailed texture information, thereby minimizing the loss of contextual information. The two branches effectively suppress the complex background and clutter interference by synergistically exploiting the interaction of local and non-local features to achieve accurate extraction of multi-scale features. Next, by utilizing the Euclidean distance and combining position and shape information, we propose a shape-position similarity metric to solve the problem of position deviation sensitivity in small target-dense scenes, thereby balancing the allocation of positive and negative samples. In a comprehensive comparison with 11 detectors from one-stage, two-stage, and DETR series on the SSDD and HRSID datasets, our method achieves mAP scores of 68.8% and 98.3%, and mAP50 scores of 70.8% and 93.8%, respectively. In addition, our model is highly efficient, with just 2.4 M parameters and a computational load of only 6.4 GFLOPs, outperforming the comparison methods. The proposed method shows excellent detection performance under complex backgrounds and ship targets of different scales. While reducing the false detection rate and missed detection rate, it has a low model parameter amount and computational complexity.

Electrowetting electronic paper employs a subtractive color mixing system for color display, which results in a smaller color gamut and potential color distortion. Additionally, it relies on ambient light's diffuse reflection, leading to insufficient brightness. To address these issues, this paper proposes a color space transformation and image adaptive enhancement algorithm for color electrowetting. The algorithm converts the image from the RGB color space to the HSV space, using CLAHE to evenly distribute saturation and improve color performance. The luminance channel is enhanced through guided filtering combined with an improved Retinex algorithm, preserving detail and edge information, and ensuring the electrowetting display maintains realistic visual effects under the same lighting conditions. Experimental results show that the algorithm improves PSNR, SSIM, FSIM, and FSIMc by 19%, 10.8%, 19.19%, and 19.54%, respectively. This algorithm significantly enhances the display performance of electrowetting electronic paper, laying a solid foundation for its commercialization.

This paper proposes a smartphone image quality assessment method that combines the Swin-AK Transformer based on alterable kernel convolution and manual features based on dual attention cross-fusion. Firstly, manual features that affected image quality were extracted. These features could capture subtle visual changes in images. Secondly, the Swin-AK Transformer was presented and it could improve the extraction and processing of local information. In addition, a dual attention cross-fusion module was designed, integrating spatial attention and channel attention mechanisms to fuse manual features with deep features. Experimental results show that the Pearson correlation coefficients on the SPAQ and LIVE-C datasets reached 0.932 and 0.885, respectively, while the Spearman rank-order correlation coefficients reached 0.929 and 0.858, respectively. These results demonstrate that the proposed method in this paper can effectively predict the quality of smartphone images.

Due to the low high air pressure during the flight, if a fire occurs in the cargo hold of the aircraft, the smoke particles are suspended in mid-air. The traditional smoke detector is difficult to detect, and there is also a high false alarm rate and difficult visualization in other environments, an image-based fire detector was designed, and the improved YOLOv5s algorithm was used to realize the pyrotechnic target detection. First, the backbone network is replaced with a lightweight GhostNet backbone network to facilitate hardware deployment. A collaborative attention module is embedded in the connection between the backbone and the converged network to strengthen the extraction of effective features. Then, according to the development and change characteristics of fire targets, the C3 structure in the feature fusion network was improved, the VoV-GSCSP module was built, and the Slim-ASFF module was embedded between the fusion network and the detection head, so as to jointly strengthen the feature fusion of different scales and realize the further lightweight of the overall network. Finally, the regression loss is replaced by focal EIOU, which solves the problem of penalty term failure and improves the prediction ability of positive samples. The image-based aviation fire detector takes the domestic AI chip RK3588 as the core, connects to the CMOS image sensor for data collection, and realizes information interaction with the airborne display system through the network. The test results show that the equipment can be arranged at the top four corners of the cargo compartment of the simulated aircraft, which can realize the flame alarm within 10 seconds and the smoke alarm within 20 seconds, which provides a feasible solution for ensuring the safety of the aircraft.

To address the low efficiency in metal surface defect detection, and the problems related to numerous model parameters and low precision, a lightweight detection method based on an improved YOLOv8n was proposed. The partially inverted bottleneck cross-stage partial fusion (PIC2f) module was introduced, replacing the bottleneck module with a partial IRMB bottleneck (PIBN) module. This combination of partial convolution and inverted residual blocks reduced the algorithm’s parameters and enhanced the model’s feature extraction ability. The attention-based intra-scale feature interaction (AIFI) module was applied, integrating location embedding and multi-head attention to improve the model’s small-target detection performance. Lastly, the average pooling down sampling (ADown) module replaced traditional convolution as the feature reduction module, reducing parameters and computational complexity while maintaining detection accuracy. The experimental results show that, compared to YOLOv8n, the PIC2f-YOLO method improves mAP50 by 2.7% on the NEU-DET steel defect dataset and reduces parameters by 0.403 M. Generalization experiments on aluminum sheet surface industrial defects, PASCAL VOC2012 and surface defects of strip alloy functional material datasets also confirm the method’s effectiveness.

A dynamic SAR image target detection algorithm integrating space-frequency domains is proposed to address challenges such as significant feature differences in SAR image samples, imbalanced target scales, and high speckle noise in the background, which result in low detection accuracy and slow inference speed. First, a dual-stream perception strategy constructs spatial-frequency perception units, leveraging dynamic receptive fields and fractional-order Gabor transforms to enhance the model’s ability to capture spatial diversity and frequency scattering features. This way improves the retention of global contextual information, accelerates inference, reduces the similarity of feature mapping patterns, and mitigates background noise interference, effectively reducing missed and false detections. Second, a re-parameterization-based adaptive feature fusion module is designed to optimize interactions across multi-scale features, enriching feature diversity, alleviating mapping discrepancies and information loss caused by feature sampling, and enhancing the salience of small target and key frequency information during fusion, thereby improving detection precision. Finally, the DY_IoU dynamic regression loss function is introduced, utilizing adaptive scale penalty factors and a dynamic non-monotonic attention mechanism to address anchor box expansion and positional deviation, further enhancing the localization and detection capabilities for multi-scale targets. This way also accelerates model convergence and reduces computational overhead. Experiments conducted on the public datasets SAR-Acraft-1.0 and HRSID demonstrate that the proposed method achieves mAP@0.5 values of 95.9% and 98.8%, respectively, representing 5.2% and 1.2% improvements over baseline models and outperforming other comparison algorithms. These results indicate that the proposed algorithm improves detection accuracy and exhibits strong robustness and generalization capabilities.

To address issues of insufficient receptive fields and weak connections between global and local features in unsupervised domain adaptive person re-identification, a multi-scale feature interaction method was proposed. Firstly, the feature squeeze attention mechanism compressed image features, which were then fed into the network to enhance rich local information representation. Secondly, the residual feature interaction module encoded global information into the features by interaction, while increasing the model's receptive field and enhancing its ability to extract pedestrian features. Finally, a bottleneck module based on partial convolution conducted convolution operations on the part of the input channels, reducing redundant computations and improving spatial feature extraction efficiency. Experimental results on three adaptation datasets demonstrate that the method mAP reached 82.9%, 68.7%, and 26.6%, the Rank-1 reached 93.7%, 82.7%, and 54.7%, the Rank-5 reached 97.4%, 89.9%, and 67.5%, by comparison with baseline, respectively, demonstrating that the proposed method allows for better pedestrian features representation and improved recognition accuracy.

Aiming at the issues of discontinuous road edge segmentation, low accuracy in segmenting small-scale roads, and misclassification of target roads in high-resolution remote sensing imagery, this paper proposes a road extraction method that integrates ResNeSt and multi-scale feature fusion for road extraction from remote sensing imagery. Referencing the ResNeSt network module, a U-shaped network encoder is constructed to enable the initial encoder to extract information more entirely and ensure more continuous segmentation of target edges. Firstly, Triplet Attention is introduced into the encoder to suppress useless feature information. Secondly, convolutional blocks replace max pooling operations, increasing feature dimensionality and network depth while reducing the loss of road information. Finally, a multi-scale feature fusion (MSFF) module is utilized at the bridge connection between the encoder and decoder networks to capture long-range dependencies between regions and improve road segmentation performance. The experiments were conducted on the Massachusetts Roads dataset and the DeepGlobe dataset. The experimental results demonstrate that our proposed method achieved Intersection over Union scores of 65.39% and 65.45%, respectively, on these datasets, representing improvements of 1.42% and 1.74% compared to the original MINet model. These findings indicate that the ResT-UNet network effectively enhances the extraction accuracy of road features in remote sensing imagery, providing a novel approach for interpreting semantic information in remote sensing images.

A lightweight Swin Transformer and multi-scale feature fusion (EMA) module combination is proposed for face expression recognition, which addresses the problems of the Swin Transformer model, such as excessive parameter quantity, poor real-time performance, and limited ability to capture the complex and small expression change features present in the expressions. The method first uses the proposed SPST module to replace the Swin Transformer block module in the fourth stage of the original Swin Transformer model to reduce the number of parameters of the model and realize the lightweight model. Then, the multi-scale feature fusion (EMA) module is embedded after the second stage of the lightweight model, which effectively improves the model's ability to capture the details of facial expressions through multi-scale feature extraction and cross-space information aggregation, thus improving the accuracy and robustness of facial expression recognition. The experimental results show that the proposed method achieves 97.56%, 86.46%, 87.29%, and 70.11% recognition accuracy on four public datasets, namely, JAFFE, FERPLUS, RAF-DB, and FANE, respectively. Compared with the original Swin Transformer model, the number of parameters of the improved model is decreased by 15.8% and the FPS is improved by 9.6%, which significantly enhances the real-time performance of the model while keeping the number of parameters of the model low.

A fiber microprobe vibration measurement system based on internal modulation has been developed. A sinusoidal phase modulated laser source is generated by modulating the current of the distributed feedback laser (DFB) with a sinusoidal signal. After passing through a fiber circulator and a fiber microprobe, the output laser beam is used to measure the displacement of a vibration source. The returned laser beam interferes with the reference laser reflected by the fiber microprobe to generate a phase generated carrier (PGC) interference signal. A real-time PGC signal processing algorithm is designed through a programmable logic gate array (FPGA) digital computing platform. A five-parameter ellipse fitting method is unitized to extract the error items introduced by additional intensity modulation and other factors and compensate for the phase nonlinear error. The fast Fourier transform (FFT) algorithm is unitized to analyze the vibration displacement. Theoretical analysis was conducted and a vibration measurement system was built. A series of experiments were conducted, including PGC signal demodulation, displacement measurement, and vibration measurement. The experimental results show that the vibration frequency range of the system covers 1142 Hz. In the 10 μm step displacement experiment, the average deviation measured is 0.173 μm. The resolution of vibration measurement is 1.221 Hz, and the harmonic distortion is less than 1.36%. The measurement system is expected to be applied in the field of precise vibration measurement.

This paper presents a low-loss fusion splice method between the nested hollow-core anti-resonant fiber (HC-ARF) and single-mode fiber (SMF) by introducing a graded-index multi-mode fiber (GIMF) as a transition fiber. The mode field matching between the nested HC-ARF and the SMF is achieved by using the GIMF as the mode field adapting fiber and expanding the mode field in the SMF by using its self-imaging effect. The effects of discharge time and discharge power on fusion splice loss during fusion splicing are explored in the experiments. Based on an optimized fusion splicing scheme, the integrity of the microstructure of the nested HC-ARF fusion splicing end face is effectively protected, and the average fusion splicing loss is as low as 0.60 dB. The experimental results provide a reference to improve the compatibility of the nested hollow-core anti-resonant fibers with the existing fiber system.

To address the issues of missed and false detections caused by significant scale differences of foreground targets, uneven sample spatial distribution, and high background redundancy in UAV aerial images, an adaptive foreground-focused UAV aerial image target detection algorithm is proposed. A panoramic feature refinement classification layer is constructed to enhance the algorithm's focusing capability and improve the representation quality of foreground sample features through the re-parameterization spatial pixel variance method and shuffling operation. An adaptive dual-dimensional feature sampling unit is designed using a separate-learn-merge strategy to strengthen the algorithm's ability to extract foreground focus features and retain background detail information, thereby improving false detection situations and accelerating inference speed. A multi-path information integration module is constructed by combining a multi-branch structure and a broadcast self-attention mechanism to solve the ambiguity mapping problem caused by downsampling, optimize feature interaction and integration, enhance the algorithm's ability to recognize and locate multi-scale targets, and reduce model computational load. An adaptive foreground-focused detection head is introduced, which employs a dynamic focusing mechanism to enhance foreground target detection accuracy and suppress background interference. Experiments on the public datasets VisDrone2019 and VisDrone2021 show that the proposed method achieves mAP@0.5 values of 45.1% and 43.1%, respectively, improving by 6.6% and 5.7% compared to the baseline model, and outperforming other comparison algorithms. These results demonstrate that the proposed algorithm significantly improves detection accuracy and possesses good generalizability and real-time performance.

This paper presents a new demodulation approach for optical fiber temperature sensors based on GaAs, leveraging reference filtering and a cross-correlation algorithm. It preprocesses the data through double Gaussian filtering for smoothing and implements an enhanced cross-correlation algorithm adopting a long-pass filter (LPF) waveform as the reference signal to demodulate the GaAs optical fiber temperature sensor. Using the correlated data from cross-correlation operations, it applies a multiple polynomial fitting strategy to further augment the precision of the cross-correlation algorithm’s demodulation. Across a temperature sensing range of ?30 to 250 ℃, the wavelength demodulation error of this method can reach ±0.0016 nm, and the temperature demodulation accuracy is ±0.388 ℃. Relative to the prevailing normalized optical intensity demodulation method, the cross-correlation algorithm employing an LPF waveform as the reference demonstrates a 2.64-fold increase in noise immunity and a 2.08-fold improvement over cross-correlation algorithms without the LPF reference waveform.

To enhance image quality in low-light conditions, an unsupervised dual-path low-light image enhancement algorithm is proposed, integrating color correction and structural information. The algorithm utilizes a generative adversarial network (GAN) with a generator that employs a dual-branch architecture to concurrently handle color and structural details, resulting in natural color restoration and clear texture details. A spatial-discriminative block (SDB) is introduced in the discriminator to improve its judgment capability, leading to more realistic image generation. An illumination-guided color correction block (IGCB) uses illumination features to mitigate noise and artifacts in low-light environments. The selective kernel channel fusion (SKCF) and convolution attention block (CAB) modules enhance the semantic and local details of the image. Experimental results show that the algorithm outperforms classical methods on the LOL and LSRW datasets, achieving PSNR and SSIM scores of 19.89 and 0.672, respectively, on the LOLv1 dataset, and 20.08 and 0.693 on the LOLv2 dataset. Practical applications confirm its effectiveness in restoring brightness, contrast, and color in low-light images.

Light field images provide users with a more comprehensive and realistic visual experience by recording information from multiple viewpoints. However, distortions introduced during the acquisition and visualization process can severely impact their visual quality. Therefore, effectively evaluating the quality of light field images is a significant challenge. This paper proposes a no-reference light field image quality assessment method based on deep learning, combining spatial-angular features and epipolar plane information. Firstly, a spatial-angular feature extraction network is constructed to capture multi-scale semantic information through multi-level connections, and a multi-scale fusion approach is employed to achieve effective dual-feature extraction. Secondly, a bidirectional epipolar plane image feature learning network is proposed to effectively assess the angular consistency of light field images. Finally, image quality scores are output through cross-feature fusion and linear regression. Comparative experimental results on three common datasets indicate that the proposed method significantly outperforms classical 2D image and light field image quality assessment methods, with a higher consistency with subjective evaluation results.

Current visible-infrared person re-identification research focuses on extracting modal shared saliency features through the attention mechanism to minimize modal differences. However, these methods only focus on the most salient features of pedestrians, and cannot make full use of modal information. To solve this problem, a quadrupl-stream input-guided feature complementary network (QFCNet) is proposed in this paper. Firstly, a quadrupl-stream feature extraction and fusion module is designed in the mode-specific feature extraction stage. By adding two data enhancement inputs, the color differences between modalities are alleviated, the semantic information of the modalities is enriched and the multi-dimensional feature fusion is further promoted. Secondly, a sub-salient feature complementation module is designed to supplement the pedestrian detail information ignored by the attention mechanism in the global feature through the inversion operation, to strengthen the pedestrian discriminative features. The experimental results on two public datasets SYSU-MM01 and RegDB show the superiority of this method. In the full search mode of SYSU-MM01, the rank-1 and mAP values reach 76.12% and 71.51%, respectively.

A dual channel encrypted free-space optical communication system based on compressive sensing and tilted fiber grating is proposed. This approach not only greatly reduces the data acquisition volume, but also enhances the security of the system since the data transmitted in the free-space is encrypted. Besides, our proposal adopts low-bandwidth and low-cost photodetectors and analog-to-digital convertors, decreasing the data acquisition volume and the cost of data transmission. Also, the approach utilizes the tilted fiber grating with a 45° tilted angle as the free-space light emitter, free-space light lateral diffraction device, and polarization-sensitive device, simultaneously. The utilization of 45° tilted fiber grating greatly enhances the systematic integration, reduces the volume of the system and improves the energy efficiency of the system. A demonstration shows that two 1 GHz and 3 GHz sinusoidal signals are employed for the 3.9 m free-space data transmission with data compression ratios of 16% and 8% achieved both in the time domain and frequency domain.

When designing metasurface systems, the actual efficiency of the metasurface is much different from the theoretical design efficiency. This can lead to stray light caused by insufficient modulation efficiency of the metasurfaces, which acts as background noise and is magnified in cascaded metasurface systems step by step, affecting system functionality. To reduce the impact of metasurfaces with limited efficiency on system performance, this paper proposes a design method for an orbital angular momentum demultiplexing system based on off-axis cascaded metasurfaces. By incorporating an off-axis phase design, the stray light generated by the reduced efficiency of the metasurface in a cascaded metasurface system is effectively eliminated. Using FDTD (finite difference time domain) simulation software for calculation and validation, the results demonstrate that the off-axis cascaded metasurface system can effectively reduce stray light caused by insufficient modulation efficiency. Compared to the on-axis system, it achieves a maximum reduction in crosstalk of 4.15 dB and an average of 80% stray light elimination, showing a significant performance advantage.

LiDAR currently mainly uses a Dammann grating as the laser beamsplitter. However, as a periodic diffraction optical device, the Dammann grating satisfies the grating equation requiring each diffraction angle's sine value to form an arithmetic progression, which cannot achieve uniform angular beam-splitting. The theoretical diffraction efficiency is also limited. This paper uses the angular spectrum and random search optimization algorithm to design a more flexible non-periodic beamsplitter. Simulations show that the metasurface beamsplitter can achieve a 70-degree field angle of 41 beams with an equal diffraction angle interval. The simulated diffraction efficiency reaches 84% which is higher than the diffraction limit of a binary phase device. In experiments, the metasurface beamsplitter has good beam-splitting uniformity and can promote the development of LiDAR.

When a vector optical field acts on the metasurface-based diffractive light sail, the maximum acceleration, self-stabilizing thrust, and attitude controllability of the diffractive light sail can be enhanced. In a vacuum environment, it is important to measure the optical force acting on the diffraction light sail to establish a comprehensive space dynamics model under the influence of vector optical fields. Based on the weak force measurement technique, we have designed an optical force measurement torsion pendulum for both regular and irregular shaped diffraction light sails. The measurement accuracy of regular-shaped light sails can be enhanced by ensuring that the size of the torsion pendulum and relative position errors of each component are strictly controlled. The force measurement has a relative error of 0.55‰. We have also designed a torsion pendulum to measure the optical force of the irregular-shaped light sails, which can hardly calculate the moment of inertia. There are two standard spheres on the torsion pendulum that can be placed or removed at any time. The magnitude of the optical force acting on the complex object can be measured by calculating the moment of inertia of the spheres. This research enhances the efficiency and flexibility of optical force measurement experiments, providing data support for applications such as laser-driven light sail and space debris remediation.

The echelle grating spectrometer has cross-dispersion characteristics, and two-dimensional spectral map reduction is the key link to determine its wavelength measurement accuracy, but the changes of spot coordinates caused by environmental changes, processing and mounting have a serious impact on the accuracy of spectral map reduction. In this paper, a spectrum reduction algorithm based on least-squares image coordinate correction is proposed for the middle-step spectrometer. Firstly, the center-of-mass coordinates of the multi-wavelength spot of the calibrated mercury lamp light source are extracted, and the coefficient matrix is constructed by using the theoretical and actual image point coordinates. The translation, scaling, and rotating coefficients of the two-dimensional image plane are obtained by the least-squares estimation method, and then a polynomial fitting is adopted to reduce the influence of residuals, to achieve the correction of the image coordinates of the spot at different wavelengths, and then achieve the accurate resolution of wavelength, and realize the wavelength accurate solving. The experimental results show that the algorithm can effectively improve the spectral image reduction accuracy of the middle-step spectrometer, and the deviation of the corrected coordinates from the ideal coordinates is less than 0.6 image element under the condition of simulating larger mounting errors, which proves that the algorithm has high accuracy.

Most attention mechanisms, while enhancing image features, do not consider the impact of local feature interaction on overall feature representation. To address this issue, this paper proposes a global pooling residual classification network guided by local attention (MSLENet). The baseline network for MSLENet was ResNet34. First, the initial layer structure was modified to retain important image information. Second, a multiple segmentation local enhancement attention mechanism (MSLE) module was introduced. The MSLE module first segmented the image into multiple small images, then enhanced the local features of each small image, and finally integrated these important local features into the global features through feature group interaction. Lastly, a pooling residual (PR) module was proposed to address the information loss problem in the ResNet residual structure and improve the information utilization between layers. The experimental results show that by enhancing the interaction of local features, MSLENet achieves good performance on multiple datasets and effectively improves the expressive ability of the network.

Optical coherence tomography (OCT) is widely used in ophthalmic diagnosis and adjuvant therapy, but its imaging quality is inevitably affected by speckle noise and motion artifacts. This article proposes a multi teacher knowledge distillation network MK-OCT for OCT super-resolution tasks, which uses teacher networks with different advantages to train balanced, lightweight, and efficient student networks. The use of efficient channel distillation method ECD in MK-OCT also enables the model to better preserve the texture information of retinal images, meeting clinical needs. The experimental results show that compared with classical super-resolution networks, the model proposed in this paper performs well in both reconstruction accuracy and perceptual quality, with smaller model size and less computational complexity.

Based on the characteristics of low light energy loss, relatively simple assembly, relatively easy debugging, and high system resolution, the critical angle focusing technology can be introduced to amplify the extracted focusing signal, improve the detection sensitivity and the smoothness of the focusing curve, and at the same time achieve a larger focal plane detection range, reducing the focusing error affected by the unevenness of the substrate itself. This paper first starts from the basic principle of the differential critical angle focusing technology and obtains the relationship between the defocus amount and the defocus signal through the Fresnel formula and the Newton formula. Secondly, a differential critical angle focusing verification system is built, a four-quadrant photodetector collects the defocus signal of a single critical angle prism, and the defocus signals received by two vertically placed four-quadrant photodetectors are differentially calculated to obtain the relationship between the differential K value and the defocus amount z. The experimental results show that when a wavelength of 532 nm laser and a numerical aperture of 0.3 projection objective are used, the focusing linear range can reach 22 μm. When a numerical aperture of 0.45 projection objective is used, the focusing linear range can reach 14 μm, and the resolution of the differential critical angle focusing method can reach 25 nm.

To address the challenges of complex background interference, multi-scale differences in targets, and the difficulty in extracting small targets from remote sensing images, this paper proposes a remote sensing image detection algorithm based on the YOLOv7-tiny model that integrates the visual center mechanism and parallel patch perception. Firstly, the algorithm introduces an explicit visual center mechanism to establish long-distance dependencies between pixels, enriching the overall semantic information of the image and improving the extraction performance of target textures. Secondly, it improves the parallel patch perception module by adjusting the feature extraction receptive fields to adapt to different target scales. Thirdly, a multi-scale feature fusion module is designed to efficiently fuse multi-layer features, thereby improving the model's inference speed. Experimental results on the RSOD dataset show that the proposed algorithm achieves improvements over YOLOv7-tiny in terms of precision, recall, and mean average precision by 1.5%, 2.4%, and 2.4%, respectively. Additionally, validation on the NWPU VHR-10 and DOTA datasets confirms the strong generalization performance of the proposed algorithm. Comparative analysis with other algorithms further demonstrates the superior performance of the proposed approach.

Focusing evaluation is the key to extending the depth of field in microscopy with stacked focus. To accurately and quickly obtain the pixel focusing position of the stacked focus image sequences and generate high-quality all-in-focus images, a focusing evaluation algorithm based on color vector space is proposed. This algorithm directly calculates color image gradients in the RGB vector space, fully utilizing the correlation between color channels, avoiding the information loss caused by traditional focus evaluation algorithms when converting color images into grayscale images, and has higher accuracy compared to simple stacking of color component gradients; Using the average Manhattan distance between the center pixel and neighboring pixels in RGB space as the focus evaluation weight can enhance the sensitivity of the focusing part, reduce the evaluation value of the defocused part, and make the focus evaluation curve characteristics tend to be idealized. Seven focusing evaluation algorithms in spatial domain, frequency domain, and statistics were selected for performance comparison experiments with the proposed algorithm. The results indicate that the proposed algorithm has better sensitivity, focusing resolution, and noise resistance in simulated and real microscopic images. The curve characteristics were significantly improved, and its application in microscope depth extension can further improve the quality of stacked focal large-depth imaging.

In order to realize the application requirements of ultra-narrowband filters in the field of optical communication and optical sensing, this paper proposes a design concept of introducing a coupled guided mode resonance in an asymmetric grating waveguide structure to complete the efficient transmission filtering for specific wavelengths. The filtering structure consists of two added subwavelength gratings stacked on a silicon-based waveguide with the same period but different filling factors. Light waves are incident vertically from the top of the composite gratings, and the asymmetric resonant coupling of multiple waveguide modes can be excited by adjusting the thickness and filling factor of the bottom grating. Numerical simulations show that extremely strong electric field enhancement can be generated using this symmetry-broken guided-mode resonant coupling. Under the premise of satisfying the high sideband rejection ratio, the structure not only realizes the ultra-narrowband filtering effect of 0.005 nm, but also has a high transmission efficiency of 99%.

In order to reduce the amount of waste metal produced by nuclear power plants, a new technology for clean control or reuse of radioactively contaminated metal components is been studied. In this study, laser composite decontamination technology was proposed for corroded 4140 steel, and compared with dry ice decontamination and laser decontamination alone. By testing the microstructure, composition distribution, metallographic structure, and three-dimensional morphology of the matrix sample, the variation trend of surface roughness and microhardness was analyzed to determine the composite decontamination effect. Finally, the engineering verification was realized in the nuclear power plant. The test and verification results show that the laser composite decontamination technology can completely remove the rust layer on the surface of 4140 steel and obtain the best decontamination effect. Furthermore, it does not affect the composition, microstructure, and mechanical properties of the matrix material. The average amount of sewage on the surface of the contaminated motor impeller after decontamination is less than 0.4 Bq/cm2, and the surface contact dose rate is less than 40 nSv/h, which meets the clearance level of solid waste in nuclear power plants. Thus, laser composite decontamination technology can provide a new method for decontamination of radioactively contaminated metal components in nuclear power plants and minimize waste.

To address the challenges faced by drones during UAV (unmanned aerial vehicle) photography in adverse conditions, such as low image recognition, obstruction by obstacles, and significant feature loss, a novel algorithm named SSG-YOLOv7 was proposed to enhance object detection from the perspective of drones in complex environments. Firstly, 12803 images were augmented from the VisDrone2019 dataset, and 1320 images were augmented from the RSOD dataset to simulate five different environments. Subsequently, anchor box sizes suitable for the datasets were clustered. The 3D non-local attention mechanism SimAM was integrated into the backbone network and feature extraction module to enhance the model's learning capabilities. Furthermore, the feature extraction module SPPCSPC was restructured to integrate information extracted from channels with different pool sizes and introduce the lightweight convolution module GhostConv, thereby improving the precision of dense multi-scale object detection without increasing the model's parameter count. Finally, Soft NMS was employed to optimize the confidence of anchor boxes, reducing false positives and missed detections. Experimental results demonstrate that SSG-YOLOv7 exhibits superior detection performance in complex environments, with performance metrics VisDrone_mAP@0.5 and RSOD_mAP@0.5 showing improvements of 10.45% and 2.67%, respectively, compared to YOLOv7.

To sparse semantics and enhance attention to key features, enhance the correlation between spatial and local features, and constrain the spatial position of features, this paper proposes a sparse feature image classification network with spatial position correction (SSCNet) for spatial position correction. This network is based on the ResNet-34 residual network. Firstly, a sparse semantic enhanced feature (SSEF) module is proposed, which combines depthwise separable convolution (DSC) and SE to enhance feature extraction ability while maintaining the integrity of spatial information; Then, the spatial position correction symmetric attention mechanism (SPCS) is proposed. SPCS adds the symmetric global coordinate attention mechanism to specific positions in the network, which can strengthen the spatial relationships between features, constrain and correct the spatial positions of features, and enhance the network's perception of global detailed features; Finally, the average pooling module (APM) is proposed and applied to each residual branch of the network, enabling the network to more effectively capture global feature information, enhance feature translation invariance, delay network overfitting, and improve network generalization ability. In the CIFAR-10, CIFAR-100, SVHN, Imagenette, and Imagewood datasets, SSCNet has shown varying degrees of improvement in classification accuracy compared to other high-performance networks, proving that SSCNet can better extract local detail information while balancing global information, with high classification accuracy and strong generalization performance.

Given that small targets are predominant in the steel surface defect areas, most existing methods cannot balance the trade-off between detection accuracy and speed. In this paper, we propose a steel surface defect detection algorithm based on YOLOv7-tiny (GBS-YOLOv7t). Firstly, we design the GAC-FPN network to fully integrate the target semantic information progressively and across layers, aiming to address the limited information flow issue in traditional feature pyramids. Secondly, we embed a bi-level routing attention (BRA) module to endow the model with dynamic query and sparse perception capabilities, thus enhancing the detection accuracy of small targets. Thirdly, we introduce the SIoU loss function to improve the training and inference capabilities of the model, and to enhance the network robustness. Experimental validation on the public dataset NEU-DET demonstrates an mAP of 72.9% and a precision of 69.9% for GBS-YOLOv7t, achieving improvements of 4.2% and 8.5%, respectively, over the original YOLOv7-tiny model. The FPS reaches 104.1 frames, indicating strong real-time performance. Compared to other detection algorithms, GBS-YOLOv7t is more effective in detecting small targets in steel surface areas, with experimental results showing that the improved algorithm better balances the detection accuracy and speed.

An improved spatio-temporal graph convolutional network for video anomaly detection is proposed to accurately capture the spatio-temporal interactions of objects in anomalous events. The graph convolutional network integrates conditional random fields, effectively modeling the interactions between spatio-temporal features across frames and capturing their contextual relationship by exploiting inter-frame feature correlations. Based on this, a spatial similarity graph and a temporal dependency graph are constructed with video segments as nodes, facilitating the adaptive fusion of the two to learn video spatio-temporal features, thus improving the detection accuracy. Experiments were conducted on three video anomaly event datasets, UCSD Ped2, ShanghaiTech, and IITB-Corridor, yielding frame-level AUC values of 97.7%, 90.4%, and 86.0%, respectively, and achieving accuracy rates of 96.5%, 88.6%, and 88.0%, respectively.

This paper proposes a novel method of compensating for the fiber optic gyroscope (FOG) temperature drift at full temperatures: the temperature field inside the NFS is constructed by multiple temperature variables, which are composed of the thermometer information built in the three inertial sensors, and then the support vector regression (SVR) is used to describe the relationship between the multiple temperature variables and the temperature drift error of the FOG, and finally the sparrow search algorithm (SSA) is applied to tune the model parameters to improve the accuracy and generalization capability. The experimental results validate the effectiveness of the proposed method, and we improve the accuracy of the NFS start-up stage from 0.0209° to 0.0101°. The performance is closely comparable to that of the stable stage, and improves the fast response capability of NFS at different initial temperatures.

In response to the problem of the large parameter size of semantic segmentation networks, making it difficult to deploy on memory-constrained edge devices, a lightweight real-time semantic segmentation algorithm is proposed based on BiLevelNet. Firstly, dilated convolutions are employed to augment the receptive field, and feature reuse strategies are integrated to enhance the network's region awareness. Next, a two-stage PBRA (Partial Bi-Level Route Attention) mechanism is incorporated to establish dependencies between distant objects, thereby augmenting the network's global perception capability. Finally, the FADE operator is introduced to combine shallow features to improve the effectiveness of image upsampling. Experimental results show that, at an input image resolution of 512×1024, the proposed network achieves an average Intersection over Union (IoU) of 75.1% on the Cityscapes dataset at a speed of 121 frames per second, with a model size of only 0.7 M. Additionally, at an input image resolution of 360×480, the network achieves an average IoU of 68.2% on the CamVid dataset. Compared with other real-time semantic segmentation methods, this network achieves a balance between speed and accuracy, meeting the real-time requirements for applications like autonomous driving.

An improved YOLOv5s network for defects detection for the cable surface of cable-stayed bridge fast and accurately is proposed. This overcomes the problems of low efficiency and poor safety of manual inspection, slow and inaccuracy of existing target detection methods because of the interference of dirt leading to wrong and missed detections. The TRANS module is added to the backbone network of conventional YOLOv5s to obtain more features of a single image and improve defect detection accuracy. Moreover, the CSP module of the neck network is replaced by the GhostBottleneck module and ordinary convolution is replaced by depth-separable convolution to reduce parameters and improve the computational speed of the network. Furthermore, the SIOU loss function is used for suppressing the oscillation of the bounding box and improving the calculation accuracy of repeatability between the prediction and the real box, which can increase the model stability. The experiments show that mAP and FPS of improved YOLOv5s network are 94.26% and 68 frames per second, respectively. The performance is better than that of Faster-RCNN, YOLOv4, and conventional YOLOv5, and it can find the surface defect for the cable of the cable-stayed bridge accurately and timely.

Biomolecules with different chirality have different or even opposite biological and pharmacological activities. Since vibration and rotation energy levels of many biomolecules lie within the terahertz range, terahertz spectroscopy has emerged as a useful tool for biomolecular identification. Nevertheless, linearly polarized light sources are used in terahertz time-domain spectroscopy, which is ineffective for identifying chiral compounds. We theoretically constructed a circularly polarized Jones matrix using a linearly polarized Jones matrix model. We also calculated the transmission circular dichroism spectrum of the sample based on the difference in transmittance of circularly polarized light, offering a useful technique for describing various chiral compounds. The spectra of (R)-(-)-Ibuprofen and (S)-(+)-Ibuprofen were investigated using a transmission terahertz time-domain spectroscopy system, and the linear polarization biased transmittance and circular polarization transmittance of (R)-(-)-Ibuprofen and (S)-(+)-Ibuprofen were computed. Additionally, the transmittance circular dichroic spectra were calculated, and the two chiral compounds' circular dichroic values reached 0.015, successfully achieving the (R)-(-)-Ibuprofen and (S)-(+)-Ibuprofen recognition effect. This technique serves as a guide for chiral molecule identification and detection using terahertz spectroscopy technology.

Detection of floating debris in rivers is of great significance for ship autopilot and river cleaning, but the existing methods in targeting floating objects in the river with small target sizes and mutual occlusion, and less feature information lead to low detection accuracy. To address these problems, this paper proposes a small target object detection method called PAW-YOLOv7 based on YOLOv7. Firstly, in order to improve the feature expression ability of the small target network model, a small target object detection layer is constructed, and the self-attention and convolution hybrid module (ACmix) is integrated and applied to the newly constructed small target detection layer. Secondly, in order to reduce the interference of the complex background, the Omni-dimensional dynamic convolution (ODConv) is used instead of the convolution module in the neck, so as to give the network the ability to capture the global contextual information. Finally, the PConv (partial convolution) module is integrated into the backbone network to replace part of the standard convolution, while the WIoU (Wise-IoU) loss function is used to replace the CIoU. It achieves the reduction of network model computation, improves the network detection speed, and increases the focusing ability on the low-quality anchor frames, accelerating the convergence speed of the model. The experimental results show that the detection accuracy of the PAW-YOLOv7 algorithm on the FloW-Img dataset improved by the data extension technique in this paper reaches 89.7%, which is 9.8% higher than that of the original YOLOv7, the detection speed reaches 54 frames per second (FPS), and the detection accuracy on the self-built sparse floater dataset improves by 3.7% compared with that of YOLOv7. It is capable of detecting the tiny floating objects in the river channel quickly and accurately, and also has a better real-time detection performance.

Due to the different reflective properties of the diffuse and specular components in composite surface objects and the limitations imposed by the camera depth of field, defocusing of sinusoidal fringes occurs in specular imaging, leading to phase errors. To achieve the efficient and high-precision measurement of composite surface objects, this paper proposes a method for three-dimensional surface topography measurement by combining defocused binary patterns with sinusoidal fringes. Firstly, the paper partitions and calibrates the defocus level of the system based on the edge and second-order blur methods, addressing the issue of varying defocus levels of the reference surface due to the tilted placement of the camera. Then, a binary fringe phase error model is established to determine the optimal fringe width and the defocus range. Finally, defocus compensation is applied to the binary fringes in the slightly defocused region, ensuring that the captured fringes are within the optimal defocus range. Three-dimensional surface topography measurement is conducted based on this approach. Experimental results show that the proposed method reduces the error in the specular component from 0.033 mm to 0.019 mm, thereby improving the accuracy of composite surface measurement.

An FPGA-based field programmable gate array (FPGA) and wavelength-modulated tunable diode laser absorption spectroscopy (WM-TDLAS) technique have been combined to develop a programmable gate array WM-TDLAS CO2 concentration detection system. Leveraging the programmable nature of FPGA chips, a digital lock-in amplifier (DLIA) with signal acquisition and modulation, as well as harmonic demodulation functions, was designed to meet the application requirements. To validate its performance, harmonic extraction tests, Q-factor assessments, and anti-noise experiments were conducted. The results revealed a linearity of 99.99% for the target frequency extraction and a Q-factor of up to 45. In the harmonic extraction experiments for signals with different signal-to-noise ratios, the maximum relative error in the mean value was only 0.91% when the signal-to-noise ratio was 43 dB. Using a distributed feedback laser with a center wavelength of 1572 nm as the light source, covering the absorption line at 6360 cm?1, and utilizing an effective optical path of 14 m in a dense multi-pass gas absorption cell, gas concentration detection experiments were carried out. The system demonstrated a fitting linearity of 99.982% between the detected concentration and the amplitude of the second harmonic. By increasing the scanning wavelength frequency, the system was capable of obtaining concentration values in 0.1 seconds. The Allan variance results showed that the detection limit of the system was 1.86 ppm when the integration time was 44 seconds. The experimental results indicate that the developed WM-TDLAS detection system based on an FPGA array features high detection accuracy, rapid response, strong stability, and miniaturization, making it suitable for real-time concentration monitoring in practical applications.

Whole slide imaging (WSI) is the main basis for cancer diagnosis and prognosis, characterized by its large size, complex spatial relationships, and diverse styles. Due to its lack of detailed annotations, traditional computational pathology methods are difficult to handle WSI tasks. To address these challenges, this paper proposes a WSI survival prediction model based on graph neural networks, BC-GraphSurv. Specifically, we use transfer learning pre-training to extract features containing spatial relationship information and construct the pathological relationship topology of WSI. Then, the two branch structures of the improved graph attention network (GAT) and graph convolution network (GCN) are used to predict the extracted features. We combine edge attributes and global perception modules in GAT, while the GCN branch is used to supplement local details, which can achieve adaptability to WSI style differences and effectively utilize topological structures to handle spatial relationships and distinguish subtle pathological environments. Experimental results on the TCGA-BRCA dataset demonstrate BC-GraphSurv's effectiveness, achieving a C-index of 0.795—a significant improvement of 0.0409 compared to current state-of-the-art survival prediction models. This underscores its robust efficacy in addressing WSI challenges in cancer diagnosis and prognosis.

Sun glint is a significant confounding factor in passive optical remote sensing images. To mitigate this issue, a polarizer is typically incorporated in front of the remote sensor, leveraging the linear polarization characteristics of sun glint. The suppression effects depend on the relative position of the sun and the remote sensor, as well as the directions of polarizers. In this paper, we introduce a novel onboard system for the real-time computation of Sun glint polarization parameters, devised specifically for a spaceborne atmospheric correction instrument. Utilizing three channel polarization images (at 0°, 60°, and 120°) in the 670 band of the spaceborne atmospheric correction, we calculate the sun glint parameters and compared them against the 6S radiation transfer model, excluding image pixels heavily influenced by the could. The system is implemented using the V5 series Field Programmable Gate Array (FPGA) as the hardware platform, and the High-Level Synthesis Tool (HLS) as the software platform. The performance of the system is verified through a simple laboratory experiment, which demonstrates a calculation deviation within 0.5°. In terms of computational efficiency, the system processes a 25x25 pixel dataset in 19.47281 ms using a 100 MHz clock, with the highest resource utilization rate reaching 41%.

A real-time image recording system was designed to record spot wander formed by two orthogonal laser beams, and a series of laser beam atmospheric transmission field experiments were conducted to compare the spot wander and study atmospheric turbulence’s anisotropy by inversion. The experimental results show that the atmospheric turbulence near the ground is anisotropic which can be manifested as follows: 1) It is related to the wind direction. A significant difference between vertical wander and horizontal was recorded when the angle between the transmission path and the wind direction was small. When the laser transmission path is orthogonal to the wind direction, vertical wander is basically the same as the horizontal. 2) It is related to real-time temperature. Among the four wander components of two orthogonal laser beams, an uneven distribution of spot wander was found, and temperature decrease increases the intensity of the uneven distribution. Based on the comprehensive analysis of research results, this paper proposes a concept named atmospheric turbulence anisotropy intensity index A which can quantify characteristics and intensity of atmospheric turbulence’s anisotropy.

The main light sources used in the clinical treatment of urological surgery are thulium-doped laser, holmium-doped laser, and green laser via the double-frequency from neodymium-doped laser, etc. In recent years, with the improvement of the output power of blue semiconductor laser diodes, 450 nm blue light has attracted growing attention and been applied in bladder tumor resection surgery, offering advantages such as clean cutting, minimal bleeding, and no adverse coagulation of adjacent tissues. This work focuses on the solution for a high-stability fiber-coupled output blue laser source for urological surgery applications. A 350 W fiber-coupled blue semiconductor laser is built by utilizing four 100 W arrayed blue laser units as the light source. The optical field transmission characteristics of the multi-emitter array are analyzed, and the far-field distribution of optical intensity exhibits a dual-peak structure with a peak angle of arcsin(5λ/4γd). By applying the spatial beam combining technique, we have successfully achieved the cross-interference of the slow-axis beams, thereby obliterating the emission dead zone. A polarization beam combining scheme is performed to rotate the polarization state of one beam from P-polarized to S-polarized, and then combine it orthogonally with another P-polarized beam, resulting in compression of the spacing between fast-axis beams and improved beam brightness. The collimating structure reduces the divergence angles of the fast and slow axes to 0.6981 mrad and 1.0123 mrad, respectively. The fast axis is expanded by a factor of 1.2 to transform the output beam profile into a square shape. Finally, we obtain a blue laser with a power of 358 W, an output fiber of 200 μm/NA 0.22, a beam combining efficiency of 89.5%, an electro-optical conversion efficiency of 31.3%, and power fluctuation less than 0.4%. Using the water-cooled fiber to couple out the light beam, this high-power laser source may serve as an ideal medical solution for clinical treatment in urological surgery.

Aiming at the problem of low positioning accuracy of laser SLAM algorithm in indoor scenes with feature scarcity and narrow corners, a laser inertial SLAM method based on planar extension and constraint optimization is proposed. The IMU is fused in laser SLAM, and the laser point cloud is position compensated and key frames are judged according to the IMU state estimation results. The global planar map is constructed, the planar extraction of key frames is performed based on the RANSAC algorithm and combined with the pre-extraction method to track the planar features in order to reduce the time cost, and the fitting results are optimized by iPCA to remove the effect of noise on the RANSAC. Using the distance from the point to the surface to construct the plane constraint optimization equation, and integrate it with the edge point constraints and pre-integration constraints in a unified way to establish a nonlinear optimization model, and solve to get the optimized plane information and key frame bit position. Finally, to verify the effectiveness of the algorithm, experiments are carried out on the M2DGR public dataset and private dataset respectively, and the experimental results show that the present algorithm performs well on most of the public datasets, especially in the private dataset compared with the widely used fast-lio algorithm, the localization accuracy is improved by 61.9%, which demonstrates good robustness and real-time performance.

In recent years, the rapid development of LiDAR technology has played an increasingly important role in the field of scientific research and industry, giving new vitality to remote sensing, imaging, environmental perception and other fields. A special issue on "Innovation and application of LiDAR" was organized, which is around the single photon LiDAR, synthetic aperture LiDAR, FM continuous wave LiDAR, environmental factors LiDAR, imaging LiDAR, the new LiDAR precision measurement technology, as well as the recent main research progress. As a technical research reference and a exchange platform for scholars in related fields, this special issue will actively promote the research process of China's LiDAR technology.

With the rapid development of single-photon detectors and technologies, single-photon LiDAR with photon-level sensitivity has become a popular research topic. It plays an increasingly important role in fields such as remote sensing and mapping, intelligent driving, and consumer electronics. This paper focuses on LiDAR technologies and systems employing single-photon avalanche diode detectors, introducing three single-photon LiDAR detection principles: pulse accumulation, coding modulation and chirp modulation. Considering the importance of detectors and algorithms, it outlines the current development status of single-photon detectors and typical processing algorithms. It also reviews the applications and typical systems of single-photon LiDAR in long-distance detection, complex scene sensing, satellite/airborne remote sensing and mapping, intelligent driving navigation and obstacle avoidance, and 3D sensing in consumer electronics. Lastly, the paper analyzes the future development trends and forecasts the potential challenges of single-photon LiDAR technology in detectors, algorithms, systems, and application domains.

The optimization and design of optical systems is an important research direction in LiDAR. In this paper, the advantages of diffractive optical elements (DOE), such as high design degree of freedom and large dispersion, are used in the receiving end of LiDAR. The focusing and filtering effects are realized at the same time, which reduces the complexity of the optical system. Based on the principle of diffractive optical elements, the optical characteristics of diffractive optical elements are simulated and analyzed. The LiDAR ranging experiment is completed by using the diffractive optical element as the optical receiving end of the LiDAR. It proved that the diffractive optical elements have both a focusing effect and a narrow-band filtering effect. The experimental results are basically consistent with the simulation. Using the advantages of diffractive optical elements in LiDAR, the lightweight, integration, and high efficiency of LiDAR can be realized.

Synthetic aperture ladar is an advanced active optical imaging technology that overcomes the diffraction limit of the traditional optical system. This technology is inspired by the working principles of synthetic aperture radar in the microwave band. Compared with synthetic aperture radar, synthetic aperture ladar has several advantages such as fast imaging speed, high resolution, and its images being similar to what the eye is used to seeing, thanks to its operation wavelength, which makes synthetic aperture ladar potentially valuable. This paper aims to provide a comprehensive review of the progress made in key technologies related to synthetic aperture ladar, including its system model and basic theory, system design and structures, laser phase noise suppression technology, motion compensation technologies, and imaging algorithms. Furthermore, some important outdoor experiments at home and abroad were summarized. At last, the difficulties and challenges for the subsequent implementation of engineering were discussed.

LiDAR plays an important role in the remote sensing of environmental elements due to its active emission, high detection accuracy, good real-time performance, and high spatial and temporal resolution. Based on the coupling relationship between the scattering spectrum and the medium environment, the direct scattering spectrum LiDAR can invert the multiple environment elements by directly measuring the energy dimension and the spectral dimension multi-feature information such as energy, frequency shift, linewidth, etc. In this paper, the recent advances in spectrum characteristics research and spectrum detection techniques of direct scattering spectrum LiDAR are briefly summarized. The detection theory and inversion models of underwater and atmospheric direct scattering spectrum are mainly introduced, as well as the existing measurement methods of direct scattering spectrum.

Abandoned objects on the road significantly impact traffic safety. To address issues such as missed detections, false alarms, and localization difficulties encountered in detecting of small-to-medium-sized abandoned objects, this paper proposes a method for detecting and locating abandoned objects on the road using image guidance and point cloud spatial constraints. The method employs an improved YOLOv7-OD network to process image data, extracting information about two-dimensional target bounding boxes. Subsequently, these bounding boxes are projected onto the coordinate system of the LiDAR sensor to get a pyramidal region of interest (ROI). Under the spatial constraints of the point cloud within the ROI, the detection and localization results of abandoned objects on the road in three-dimensional space are obtained through a combination of point cloud clustering and point cloud generation algorithms. The experimental results show that the improved YOLOv7-OD network achieves recall and average precision rates of 85.4% and 82.0%, respectively, for medium-sized objects, representing an improvement of 6.6% and 8.0% compared to the YOLOv7. The recall and average precision rates for small-sized objects are 66.8% and 57.3%, respectively, with an increase of 5.3%. Regarding localization, for targets located 30-40 m away from the detecting vehicle, the depth localization error is 0.19 m, and the angular localization error is 0.082°, enabling the detection and localization of multi-scale abandoned objects on the road.

To address the ineffective reception of larger scanning field echoes by the small photosensitive area of the indium gallium arsenide InGaAs detector in the 1550 nm wavelength band, a receiver device suitable for near-range wide field-of-view applications has been designed. The optical system at the receiving end utilizes an afocal telecentric structure as the receiving antenna, achieving a reception field-of-view of 36° at a photosensitive area of 1 mm. The relative illuminance exceeds 95%, demonstrating excellent light collection and transmission characteristics. Additionally, the receiver circuit adopts a T-network amplification structure combined with a moment identification circuit, utilizing the TDC7200 to achieve high-precision time measurements. The flight time measurement accuracy is less than 120 ps within a range of 200 ns, and the experimental results demonstrate ranging accuracy better than 2 ns within an 8 m distance, meeting the requirements for near-range detection.

For large complex electromechanical systems such as aeronautics and astronautics, the demand for length measurement has also evolved from one-dimensional displacement measurement to multi-objective, flexible and extensible multiline parallel measurement. The frequency-modulated continuous-wave lidar length measurement technology has the advantages of anti-interference, high precision and no cooperation. In this paper, a multiline parallel coherent precision length measurement method based on frequency modulation continuous wave laser lidar is studied. To solve the problem of fiber dispersion and auxiliary fiber drift in the traditional length measurement system, a three-channel Mach-Zehnder interferometry system containing a hydrogen cyanide gas absorption cell is proposed. On this basis, the precision length multiline parallel measurement is realized based on the optical switch structure. The experimental results of precision guide rail length measurement show that the proposed method is less than 25 μm in the length range of 3.6 m. The four-way parallel length measurement is realized. The absolute error of the multi-channel measurement results is not more than 30 μm compared with the measurement results of a commercial laser interferometer.

In modern science and technology, LiDAR plays a key role in automatic navigation, industrial mapping and other fields, but the traditional phase ranging system has many problems such as low measurement accuracy and complex structure. A new high-precision phase laser radar ranging system is proposed in this paper. The system adopts the phase difference detection method of laser control with the same frequency reference, including the optical structure optimization of the laser transmitting and receiving modules, and the amplification filtering and differential mixing processing of the receiving circuit, and finally produces a high-resolution phase discrimination system based on AD8302. The experimental results show that the measurement accuracy of the system is millimeter level, which is simple and practical and can meet the needs of a wide range of practical applications. This research provides a feasible solution for LiDAR technology in high-precision distance measurement.

We demonstrate a new laser detection and ranging (LiDAR) system for reconstructing remote targets in three dimensions (3D). In this system, a probe beam, of which wavelength is 532 nm, working at Bessel mode rather than Gaussian mode, exhibits a typical intensity distribution of a bright central spot and some surrounding rings, and takes advantage of non-diffraction character in long-distance ranging. It is an attractive way to improve the imaging resolution of the LiDAR system. Combined with the Si-APD, we built a long-distance LiDAR system and completed the verification experiment. The results indicate that we can achieve an 18.1-μrad angle resolution in long-range target imaging, which provides an effective solution for high-resolution remote imaging.

Spatial self-organization is a common phenomenon in many natural ecosystems. The "fairy circle" is a typical spatial self-organization structure that has significant impacts on the ecological functions of the salt marsh vegetation ecosystems. Obtaining the spatial pattern and spatiotemporal changes of the "fairy circle" can provide important scientific support for clarifying its ecological evolution mechanism. In this study, a machine learning method based on random forest is used to intelligently identify and extract the "fairy circle" in salt marsh vegetation using the spatial-spectral information from unmanned aerial vehicle (UAV) LiDAR. First, the effects of the distance, incident angle, and specular reflection on intensity data are eliminated using the laser radar equation and the Phong model. Second, the corrected intensity data are filtered to separate the vegetation from the ground. Third, a series of spatial features and geometric variables are used to classify the normal vegetation and "fairy circles" using the random forest algorithm. The results demonstrate that the proposed method can accurately extract "fairy circles" from UAV LiDAR 3D point cloud data without requiring manual experience-based parameter settings. The overall accuracy of the proposed method is 83.9%, providing a high-precision method for the spatiotemporal distribution inversion of "fairy circles" and technical references for 3D point cloud data processing based on machine learning.

Spaceborne telescopes for gravitational wave detection crucially collimate bidirectional beams in ultra-long interferometric optical paths. The faint optical path changes due to gravitational waves demand pm-level optical path length stability and below 10?10 level backscattered light in the telescopes. The ultra-high-level specifications requirements are out of state-of-the-art testing techniques. The development of testing and evaluation techniques for spaceborne telescopes is a crucial prerequisite for the success of the space gravitational wave detection program. This paper overviews the development of spaceborne gravitational wave detection telescopes, focusing on the optical path length stability and backscattered light testing status, results, and further plans, providing reference in the testing and evaluation of Chinese spaceborne gravitational wave detection telescopes.

The detection of space gravitational waves is expected to reveal more mysteries of the universe. With the support of the National Key Research and Development Program of China, "Special issue on telescopes for space gravitational wave detection (II)" was organized by the journal Opto-Electronic Engineering. These papers in the special issue introduce the recent major research progress of the designs and analyses, construction and adjustable, and testing and evaluation of telescopes for space gravitational wave detection. They will provide a communication platform for the relevant field scholars and experts, and will actively promote the research process of the space gravitational wave detection project in China.

The spaceborne telescope plays a critical role in detecting gravitational waves in space. Given transmission distances of approximately 109 meters between different constellations, there are stringent requirements for nanoradian precision in telescope pointing accuracy. Accurate pointing deviation measurement and calibration are essential prerequisites for achieving high-precision pointing in space-based gravitational wave detection telescopes. To meet the ground testing and sensor calibration needs for space telescopes' pointing deviation, this paper introduces a novel high-precision measurement method based on the Hartmann principle. By utilizing the concept of multi-aperture spatial reuse, this approach mitigates the impact of various sources of random errors, significantly improving the precision of pointing deviation measurements. The paper conducts an analysis and optimization of Hartmann sensor parameters, establishing a quantitative relationship between sensor parameters and pointing deviation measurement accuracy. The research findings demonstrate that the multi-aperture spatial reuse method based on the Hartmann principle can achieve highly precise measurements of telescope pointing deviations, with the accuracy as low as 0.32 nrad. This method offers a feasible approach and serves as a reference for ground testing and in-orbit sensor calibration of space-based gravitational wave detection telescopes.

The space gravitational wave telescope is a key payload of gravitational wave detection satellites, responsible for both beam expansion and compression. Optical path stability is a crucial indicator for the telescope, closely related to its structural stability. To meet the stringent requirements for ultra-high optical path stability and structural stability in gravitational wave detection missions, it is necessary to investigate the measurement of structural deformations in the telescope. This paper presents a study on multi-degree of freedom deformation measurement for space gravitational wave telescopes, focusing on addressing the coupling issues in multi-degree of freedom measurement and conducting a detailed analysis of error sources. During the development phase of the space gravitational wave telescope, this measurement method is expected to meet the demands for multi-degree of freedom deformation measurement, providing data feedback on multi-degree of freedom deformations for telescope design and offering guidance for optical path stability research.

In gravitational wave telescopes, the energy of the collected space target light signals is dwarfed by the energy of stray light, necessitating robust stray light suppression for reliable telescope operation. Due to the inherent unpredictability of scattered light and the intricate nature of opto-mechanical systems, the formulation of stray light suppression strategies often involves complex mathematical modeling, substantial expertise, and iterative simulations. This paper introduces a Reinforcement Learning-based approach to devise the stray light suppression scheme within a Monte Carlo ray tracing environment, specifically for space gravitational wave telescope systems. Our empirical findings confirm the efficacy of this methodology in generating effective stray light suppression strategies, yielding favorable suppression performance. This study contributes a novel, efficient, and adaptable solution to the stray light challenges faced in space gravitational wave detection as well as other high-precision optical systems, thereby holding extensive applicative promise.

This paper mainly introduces the development of space gravitational wave transmission and laser propagation in space gravitational wave detection. We profile the calculation methods used in the simulation of laser propagation and jitter noise in space-based laser interferometry. Compared with ground detection, space gravitational wave detection can effectively reduce noise and increase the length of the interference arm to realize high-precision gravitational wave detection. Under the distance of millions of kilometers and the precision requirements of the picometer level, it is necessary to consider the phase noise caused by pointing jitter with the telescope. Research has shown that defocus and astigmatism are the main aberrations affecting jitter noise at a distance of 2.5×109 m. There is a deviation between the phase stationary point and the origin position. To minimize the phase noise, the telescope angle needs to be adjusted. The gravitational wave detection at the phase stationary point can effectively reduce the phase noise and the requirements of the telescope exit pupil wavefront RMS. The large defocus and small coma can make the phase stationary point close to the optical axis and increase the received laser power.

In order to meet the application requirements of pimi-level stability and λ/30 wavefront error in space gravitational wave detection, an optical and mechanical integrated analysis and optimization method is proposed. Firstly, the position analysis of the support points on the main mirror side and the topology optimization of the support structure were carried out. Then, based on the flexibility matrix of parallel Bipod linkage support, the evaluation function of each structure parameter is established, and the value range of flexible support size parameters was preliminarily determined by Matlab analysis. Finally, an integrated optical and mechanical simulation platform was built to further optimize the structure.The results show that the first-order frequency of the system is 392.23 Hz, and the deformation of the primary mirror surface deformation under gravity and temperature loads is better than λ/60. Under thermal disturbances in a space environment of 10 μK/Hz1/2, the dimensional stability of the primary mirror component is at a level of 10 pm/Hz1/2.

In order to achieve the measurement of gravitational wave signals in the millihertz frequency band, the space-based gravitational wave detection projects such as LISA, TianQin, and Taiji projects, which are based on laser interference systems, require the hardware noise floor of the interferometers to be lower than the interstellar weak light shot noise limit. This imposes stringent engineering specifications on the optical-mechanical design and the corresponding interferometer payload. This paper approaches the issue from the perspective of detection mode selection and derives the expressions of readout noise and stray light noise in the interference signal under the single detector mode and the balanced mode. Furthermore, a detailed discussion is provided on the weak-light interference process of the scientific interferometer. The results demonstrate that the balanced mode is capable of suppressing the interference phase noise caused by laser power fluctuations and backscattered stray light across multiple orders of magnitude. However, the suppression capability is constrained by the unequal splitting property of the beam combiner. To address this, a relative gain factor is introduced to compensate for the unequal splitting property of the beam combiner. Further analysis reveals that electronic gain compensation can only eliminate the impact of unequal splitting on one of the two noises rather than both simultaneously. Therefore, a balance must be struck in selecting gain compensation between the suppression of laser power fluctuation noise and stray light noise. Even with this consideration, the balanced mode still offers significant noise suppression capabilities at a magnitude difference, thus potentially reducing the engineering requirements for laser power fluctuations and telescope backscattered stray light.

Terahertz pulse radiation can be generated by pumping semiconductor surfaces or semiconductor heterojunctions with ultrashort laser pulses. Based on the model of terahertz generation from metal-insulator-semiconductor heterostructure pumped by ultrashort laser pulses, the changes of carrier density and velocity in semiconductors are studied through numerical simulations and theoretical analysis. The influences and physical mechanisms of ultrashort laser pulse duration, carrier lifetime, and semiconductor thickness on the terahertz generation are analyzed as well. The results show that the increase of the laser pulse duration increases the amplitude of the terahertz pulse but decreases its central frequency and bandwidth. The increases of the carrier lifetime and thickness of the semiconductor have different influences on the central frequency and bandwidth of terahertz pulse. By analyzing the influence of different parameters on the terahertz generation, the pathways and parameters of optimizing the terahertz generation are obtained. The results of this paper provide a good theoretical reference for the related experiments.

Aiming at the phenomenon that complex background often exists in non-motorized drivers' helmet detection and the diversity of detection target scales often exists in the detection scene, which in turn leads to the low detection efficiency and misdetection and omission, a YOLOv8 algorithm oriented to the detection of traffic helmets is proposed. Combining the advantages of GSConv and CBAM in the C2f module, the C2f_BC module is designed to effectively improve the feature extraction capability of the model while reducing the number of model parameters. A multi-core parallel perception network (MP-Parnet) is designed to improve the model's perception and feature extraction ability for multi-scale targets so that it can be better applied to complex scenes. To alleviate the problem of positive and negative sample imbalance in complex scenes, Focaler-IoU is introduced based on the original model's loss function CIoU, and a threshold parameter is introduced to improve the calculation of the IoU loss, thus alleviating the phenomenon of positive and negative sample imbalance,and effectively improves the model's accuracy of target frame localization in complex background. The experimental results show that compared with the original model, the improved YOLOv8n maintains a decrease in the number of parameters while the mAP50 and mAP50: 95 increase by 2.2% and 1.9% on the self-built dataset Helmet, and 1.8% and 1.9% on the open-source dataset TWHD, which suggests that the improved model can be better applied to the helmet detection of non-motorized drivers in the scenario.

To address the issues of insufficient utilization of multi-scale semantic information and high computational costs resulting from the generation of lengthy sequences in existing Transformer-based semantic segmentation networks, this paper proposes an efficient semantic segmentation backbone named MFE-Former, based on multi-scale feature enhancement. The network mainly includes the multi-scale pooling self-attention (MPSA) and the cross-spatial feed-forward network (CS-FFN). MPSA employs multi-scale pooling to downsample the feature map sequences, thereby reducing computational cost while efficiently extracting multi-scale contextual information, enhancing the Transformer’s capacity for multi-scale information modeling. CS-FFN replaces the traditional fully connected layers with simplified depth-wise convolution layers to reduce the parameters in the initial linear transformation of the feed-forward network and introduces a cross-spatial attention (CSA) to better capture different spaces interaction information, further enhancing the expressive power of the model. On the ADE20K, Cityscapes, and COCO-Stuff datasets, MFE-Former achieves mean intersection-over-union (mIoU) scores of 44.1%, 80.6%, and 38.0%, respectively. Compared to mainstream segmentation algorithms, MFE-Former demonstrates competitive segmentation accuracy at lower computational costs, effectively improving the utilization of multi-scale information and reducing computational burden.

In response to the problem of low sensitivity of fiber Bragg grating (FBG) in temperature measurement applications, this paper proposes a new solidification sensitization method based on an EpoCore adhesive-coated FBG sensor with a high thermal expansion coefficient. A FBG sensor model was constructed using COMSOL software, and the deformation of the grating area before and after sensitization was simulated. The results showed that EpoCore adhesive sensitization increased the deformation of the FBG sensor by about 10 times compared to before sensitization. Using ultraviolet exposure technology to write FBG onto thulium tin co-doped optical fibers directly and conducting in-depth research on the solidification process of EpoCore-coated FBG sensors, the resonant peak of the solidified and sensitized FBG exhibits good linearity with temperature drift when baked at 120 ℃ for 2 hours, irradiated with ultraviolet light for 5 hours, and baked at 120 ℃ for 3 hours. The temperature sensitivity is 90.45 pm/℃, about 9 times higher than the unsolidified and unsensitized FBG sensor. The sensitized FBG sensor is used to measure environment temperatures of ?20 ℃, 30 ℃, 60 ℃ and 100 ℃, and the RMSE for these temperatures is less than 1.3 ℃. Compared with traditional metal coating and polymer coating packaging methods, EpoCore adhesive has a more significant sensitization effect, which provides a new idea for the application of FBG sensors in temperature measurement.

Based on the four-dimensional representation of the two-plane model, the light field camera captures spatial and angular information of the three-dimensional scene simultaneously at the expense of image spatial resolution. To improve the spatial resolution of light field images, a two-way guided updating network for light field image super-resolution is built in this work. In the front of the network, different forms of image arrays are used as inputs, and the residual series and parallel convolution are constructed to realize the decoupling of spatial and angular information. Aiming at the decoupled spatial information and angular information, a two-way guide updating module is designed, which adopts step-by-step enhancement, fusion, and re-enhancement methods to complete the interactive guidance iterative update of spatial and angular information. Finally, the step-by-step updated angular information is sent to the simplified residual feature distillation module to realize data reconstruction. Many experimental results have confirmed that our proposed method achieves state-of-the-art performance while effectively controlling complexity.

Optic nerve injury is one of the primary causes of vision loss, so accurately assessing the extent of optic nerve fiber damage is critically important for effective treatment and rehabilitation. In this manuscript, the optic nerves of pig eyes are imaged before and after injuries using a swept-source polarization-sensitive optical coherence tomography system built with polarization-maintaining fiber components. The microstructure and birefringence characteristics of the optic nerve are observed and reflected through Stokes parameters Q, U, and V, representing the polarization state of the detection light. It was found that the V cross-sectional image has good characterization ability for the birefringence characteristics of the optic nerve. The high birefringence region and the non-birefringence or low birefringence region corresponding to the V cross-sectional image were segmented by the threshold method. The evolution of the average area and the average height of the high birefringence region in the cross-sectional image of V can reflect the damage, repair, and erosion of the optic nerve, which indicates that polarization-sensitive optical coherence tomography has a good perceptual ability for changes before and after optic nerve injury and is crucial for evaluating the degree of optic nerve injury, which can provide important reference data for early diagnosis and treatment of optic nerve injury.

To address the existing challenges of discontinuity in road region extraction and difficulty in extracting roads of different sizes, especially the misclassification of narrow roads, a novel road extraction algorithm combining element-wise multiplication and detail optimization was proposed. Firstly, an element-wise multiplication module (IEM module) was introduced in the encoder part to perform feature extraction, preserving and extracting multi-scale and multi-level road features. A Conv3×3 with a stride of 2 was used for twofold downsampling, reducing information loss during the extraction process of remote sensing images. The encoder-decoder was structured with five layers and utilized skip connections to maintain multi-scale extraction capabilities while improving road continuity. Secondly, PFAAM was employed to enhance the network's focus on road features. Finally, a fine residual network (RRN) was utilized to enhance the network's ability to extract boundary details, refining the boundary information. Experiments were conducted on the public road dataset of Massachusetts (CHN6-CUG) to test the network model, achieving evaluation metrics of OA (accuracy), IoU (intersection over union), mIoU (mean IoU), F1-score of 98.06% (97.19%)、64.52% (60.24%)、81.25% (78.66%), and 88.70% (86.85%). The experimental results demonstrated that the proposed method outperformed all the compared methods, effectively improving the accuracy of road segmentation.

This study investigates methods for suppressing dynamic aiming errors, addressing the issue of dynamic goniometer accuracy being easily affected by light intensity stability during aiming. A Gaussian spotlight intensity distribution model is established, and the mechanism of dynamic aiming error generation is analyzed. The principle of the suppression method based on the lateral photovoltaic effect is demonstrated. A dynamic aiming system is constructed, and experiments are conducted to validate the aiming repeatability and accuracy of the proposed method. The results show that the aiming repeatability is 0.19″, and the aiming accuracy is 0.15″. Compared to the light intensity threshold aiming method, the aiming error is reduced by 66%. Dynamic measurement experiments with angle blocks are conducted, applying the proposed method to the dynamic goniometer system. The results demonstrate that the system accuracy meets the calibration requirements for the grade 1 angle block gauge.

To address issues of low detection accuracy and high false-positive and false-negative rates in solar cell defect detection, this paper proposes an optimized solar cell electroluminescent (EL) defect detection model based on the YOLOv8 deep learning framework. First, a self-calibrated illumination (SCI) method is applied to preprocess low-light images, enhancing effective feature information for solar cell defects. Then, a space-to-depth (SPD) attention module is introduced, replacing the second stride convolution layer in the backbone network. This substitution avoids information loss caused by stride convolution, expands the receptive field, and reduces computational load, preserving more feature information during extraction. Next, a spatial-BiFPN (S-BFPN) network is constructed to perform multi-scale feature fusion, stabilizing defect recognition rates by addressing the shape variability of solar cell defects. Lastly, the loss function is improved by adopting MPDIoU, which resolves the issue of ineffective penalties in the original CIoU loss function. The experimental results show that the improved YOLOv8 model achieved an mAP of 96.9%, a 2.2% increase compared to the original YOLOv8. The computational load was reduced by 0.2 GFlops, and the detection speed reached a maximum of 155 f/s, demonstrating high accuracy and real-time performance, making it more suitable for industrial deployment.

To address the challenges of complex backgrounds, small targets, and dense distributions in images, an improved method called DES-YOLO is proposed. By introducing the deformable attention module (DAM), the network can dynamically focus on key regions, improving object recognition and localization accuracy. The efficient intersection over union (EIoU) loss function is employed to reduce the impact of low-quality samples, enhancing the model's generalization ability and detection accuracy. A shallow feature map layer of 160 pixel×160 pixel is added to the network head to strengthen small target feature extraction. A stepwise training strategy is also adopted to further improve model performance. Experimental results show that the mAP@50 of the model increased by 1.4% on the remote sensing dataset and by 1.7% on the textile dataset, demonstrating the broad applicability and effectiveness of DES-YOLO.

In view of the lack of spatial depth information in two-dimensional heat maps and the problems of complex calibration, high equipment cost, and limited use conditions in the existing multi-sensor schemes to establish the spatial temperature field model of the target object, this study proposes a spatial temperature field reconstruction scheme under a single thermal camera based on geometric constraints. Firstly, the temperature matrix collected by the thermal imager is remapped to the gray color space through the automatic threshold method, and the clear edge information of the target is obtained to filter out the irrelevant background. Then, the pose relationship of the target in the coordinate system of the thermal imager is directly calculated by the geometric constraints existing in the imaging process of the thermal imager and the imaging principle of the camera. Then, the 3D model of the target object is projected to the 2D heat map plane, and the texture mapping parameters of the 3D model are obtained. With the multi-view data acquisition of the target object by the thermal camera, the spatial temperature field reconstruction of the target object surface is completed. The experimental results show that, compared with the multi-sensor spatial temperature field reconstruction scheme, the average error of the single thermal imager scheme in this paper is only 4.3%, which can accurately and stably complete the spatial temperature field reconstruction of the target.

With the widespread application of low-altitude drones, real-time detection of such slow and small targets is crucial for maintaining public safety. Traditional cameras capture image frames with a fixed exposure time, which makes it challenging to adapt to changes in lighting conditions, resulting in the detection of blind spots in intense light and other scenes. Event cameras, as a new type of neuromorphic sensor, sense differences in external brightness changes pixel by pixel. They can still generate high-frequency sparse event data under complex lighting conditions. In response to the difficulty of adapting image-based detection methods to sparse and irregular data from event cameras, this paper models the two-dimensional object detection task as a semantic segmentation task in a three-dimensional spatiotemporal point cloud and proposes a drone object segmentation model based on dual-view fusion. Based on the event camera collecting accurate drone detection datasets, the experimental results show that the proposed method has the optimal detection performance while ensuring real-time performance, achieving stable detection of drone targets.

In scanning PMP, it is essential to first match different positions of the object to the same point, and then extract phase information using phase-shifting algorithms. Both pixel-matching accuracy and phase-shifting algorithms influence measurement precision. To address this, a microscopic system is employed, leveraging its telecentric optical path characteristics to achieve equal conversion between object displacement and pixel displacement. By alternately capturing white-field and fringe images, precise pixel matching is realized through optical flow in the white-field images, followed by accurate pixel matching of the fringe images based on the object's uniform motion. A set of N fringe images, closely matching a full cycle based on the initial fringe period, is selected to compute the truncated phase distribution using an arbitrary step phase-shifting method. The optimal fringe period is then identified through a probability density function, leading to the accurate extraction of phase information and the completion of the object morphology measurement. Experimental results demonstrate that the proposed method significantly enhances measurement accuracy, with the phase-shifting algorithm applying to any N≥3 images, making it particularly suitable for 3D measurements of objects in industrial production lines, achieving an RMSE measurement accuracy of about 0.008 mm.

A metamaterial-based optical solar reflector (OSR) consisting of a three-layer structure of aluminum-doped zinc oxide (AZO) metasurface, a MgF2 dielectric layer and an Ag metal reflector layer is investigated. In the thermal infrared, the AZO metasurface excites the surface equipartition excitation resonance to enhance the electromagnetic absorption, the stability of the MgF2 dielectric constant reduces the reflection caused by the absorption oscillations. In the visible light, the transparent properties of AZO and MgF2 provide the low loss for the solar radiation, and the Ag reflector layer effectively suppresses the transmission. Simulation results show that the optimized OSR has a low solar absorptivity of 17.6% in 0.3~2.5 µm and a high IR emissivity of 86.5% in 2.5~30 µm. In addition, polarization and angle of incidence have a small effect on its performance. The structure achieves good absorption in the infrared band and also has potential applications in infrared thermography, radiative cooling, and other fields.

This paper proposes a laser temperature sensor based on polarization maintaining fiber (PMF) and few-mode fiber (FMF), and conducted their experimental studies. A 20 cm polarization maintaining fiber was spliced with a 10 cm few-mode fiber and then combined with a 3 dB coupler to form a Sagnac loop, which served as the sensing probe. Light passing through the FMF excites higher-order modes. Due to the diameter mismatch between the FMF and PMF, the higher-order modes are coupled into the stress region of the PMF, exciting the cladding modes and thus improving temperature sensitivity. Experimental results show that after adding the FMF, the temperature sensitivity of the sensor increased from −0.51 nm/℃ to −0.91 nm/℃. This sensor has the advantages of high precision, easy fabrication, and intrinsic safety, making it highly promising for engineering structure safety monitoring applications.

An improved YOLOv8 algorithm is proposed to address the problems of low detection accuracy and weak generalization ability in existing roadbed slope crack detection algorithms. Firstly, a reparameterization module is embedded in the backbone network to lighten the model while capturing crack details and global information, improving detection accuracy of the model. Secondly, the C2f-GD module is designed to achieve efficient fusion of model features and enhance the generalization ability of the model. Finally, the lightweight detection head L-GNHead is designed to improve the crack detection accuracy for different scales, while the SIoU loss function is used to accelerate model convergence. The experimental results on the self-constructed roadbed slope crack dataset show that the improved algorithm improves mAP50 and mAP50-95 by 3.3% and 2.5% respectively, reduces parameters and computational costs by 46.6% and 44.4% respectively, and improves FPS by 18 frames/s compared with the original algorithm. The generalization validation results on the dataset RDD2022 show that the improved algorithm not only achieves higher detection accuracy, but also faster detection speed.

To solve the problem of small target detection in sonar images,which is difficult,low precision,and prone to misdetection and omission detection,this paper proposes an improved algorithm for small target detection in sonar images based on YOLOv8s. Firstly,considering that small targets in sonar images usually have low contrast and are easily overwhelmed by noise,an efficient multi-level screening feature pyramid network (EMS-FPN) is proposed. Secondly,since the classification branch and localization branch of the decoupled head are independent,which will increase the number of parameters of the model,and at the same time,it is difficult to effectively adapt to the detection needs of targets of different scales,resulting in poor detection of small targets,the task dynamic alignment detection head module (TDADH) is designed. Finally,to verify the effectiveness of the model in this paper,the corresponding validation was carried out on URPC2021 and SCTD expanded sonar dataset,mAP0.5 improved by 0.3% and 1.8% compared with YOLOv8s,respectively,and the number of parameters was reduced by 22.5%. The results show that the method proposed in this paper not only improves the accuracy but also significantly reduces the number of model parameters in the task of target detection in sonar images.

The dynamic depth camera has a limited single-frame field of view,and there is noise disturbance when stitching multiple frames. To deal with the aforementioned problems,a large-scale 3D target pose measurement and reconstruction method based on multi-view fusion is presented. This approach builds a hierarchical model of the depth camera's performance gradient,predicts the pose with a multi-view scanning method based on point cloud normal vectors,and fits 3D models of targets with height constraints RANSAC (height constraints RANSAC,HC-RANSAC). The depth camera installed on the end of the robotic manipulator scans and measures the target from various angles,and the sampled data is utilized to reconstruct the target model in the local coordinate system. Experimental results reveal that when compared to fixed-depth cameras and classical reconstruction approaches based on pan-tilt vision,the proposed approach has a larger reconstruction field of view and higher reconstruction accuracy. It can reconstruct huge targets at a close range,and get an excellent balance between field of vision and precision.

A new colorectal polyp image segmentation method combining polarizing self-attention and Transformer is proposed to solve the problems of traditional colorectal polyp image segmentation such as insufficient target segmentation,insufficient contrast and blurred edge details. Firstly,an improved phase sensing hybrid module is designed to dynamically capture multi-scale context information of colorectal polyp images in Transformer to make target segmentation more accurate. Secondly,the polarization self-attention mechanism is introduced into the new method to realize the self-attention enhancement of the image,so that the obtained image features can be directly used in the polyp segmentation task to improve the contrast between the lesion area and the normal tissue area. In addition,the cue-cross fusion module is used to enhance the ability to capture the geometric structure of the image in dynamic segmentation,so as to improve the edge details of the resulting image. The experimental results show that the proposed method can not only effectively improve the precision and contrast of colorectal polyp segmentation,but also overcome the problem of blurred detail in the segmentation image. The test results on the data sets CVC-ClinicDB,Kvasir,CVC-ColonDB and ETIS-LaribPolypDB show that the proposed method can achieve better segmentation results,and the Dice similarity index is 0.946,0.927,0.805 and 0.781,respectively.

To address the challenges of crowd counting in dense scenes,such as complex backgrounds and scale variations,we propose a weakly supervised crowd counting model for dense scenes,named GLCrowd,which integrates global and local attention mechanisms. First,we design a local attention module combined with deep convolution to enhance local features through context weights while leveraging feature weight sharing to capture high-frequency local information. Second,the Vision Transformer (ViT) self-attention mechanism is used to capture low-frequency global information. Finally,the global and local attention mechanisms are effectively fused,and counting is accomplished through a regression token. The model was tested on the Shanghai Tech Part A,Shanghai Tech Part B,UCF-QNRF,and UCF_CC_50 datasets,achieving MAE values of 64.884,8.958,95.523,and 209.660,and MSE values of 104.411,16.202,173.453,and 282.217,respectively. The results demonstrate that the proposed GLCrowd model exhibits strong performance in crowd counting within dense scenes.

To address the detection challenges posed by the complex backgrounds and significant variations in target scales in road damage images captured from drone aerial perspectives,a road damage detection method called MAS-YOLOv8n,incorporating a multi-branch hybrid attention mechanism,is proposed. Firstly,to address the problem of the residual structure in the YOLOv8n model being prone to interference,resulting in information loss,a multi-branch mixed attention (MBMA) mechanism is introduced. This MBMA structure is integrated into the C2f structure,strengthening the feature representation capabilities. It not only captures richer feature information but also reduces the impact of noise on the detection results. Secondly,to address the issue of poor detection performance resulting from significant variations in road damage morphologies,the TaskAlignedAssigner label assignment algorithm used in the YOLOv8n model is improved by utilizing ShapeIoU (shape-intersection over union),making it more suitable for targets with diverse shapes and further enhancing detection accuracy. Experimental evaluations of the MAS-YOLOv8n model on the China-Drone dataset of road damages captured by drones reveal that compared to the baseline YOLOv8n model,our model achieves a 3.1% increase in mean average precision (mAP) without incurring additional computational costs. To further validate the model's generalizability,tests on the RDD2022_Chinese and RDD2022_Japanese datasets also demonstrate improved accuracy. Compared to YOLOv5n,YOLOv8n,YOLOv10n,GOLD-YOLO,Faster-RCNN,TOOD,RTMDet-Tiny,and RT-DETR,our model exhibits superior detection accuracy and performance,showcasing its robust generalization capabilities.

Light field depth estimation is an important scientific problem of light field processing and applications. However,the existing studies ignore the geometric occlusion relationship among views in the light field. By analyzing the occlusion among different views,an unsupervised light field depth estimation method based on sub-light field occlusion fusion is proposed. The proposed method first adopts an effective sub-light field division mechanism to consider the depth relationship at different angular positions. Specifically,the views on the primary and secondary diagonals of the light field sub-aperture arrays are divided into four sub-light fields,i.e.,top-left,top-right,bottom-left,and bottom-right. Then,a spatial pyramid pooling feature extraction and a U-Net network are leveraged to estimate the depths of the sub-light fields. Finally,an occlusion fusion strategy is designed to fuse all sub-light field depths to obtain the final depth. This strategy assigns greater weights to the sub-light field depth with higher accuracy in the occlusion region,thus reducing the occlusion effect. In addition,a weighted spatial and an angular consistency loss are employed to constrain network training and enhance robustness. Experimental results demonstrate that the proposed method exhibits favorable performance in both quantitative metrics and qualitative comparisons.

The versatility of sensors,as a crucial part of integrated devices,is receiving increasing attention. Here,a terahertz metamaterial multifunctional sensor based on the coupling of a two-layer 3D resonant structure is introduced. The sensor consists of an upper and lower polyimide film substrate,a graphite layer attached to the lower polyimide film substrate,and a periodic double-layer 3D toothed coupling resonant structure between the graphite layer and the upper polyimide film substrate,which consists of a symmetric mountain-shaped structure in the lower layer and a symmetric concave structure in the upper layer. The three-dimensional metamaterial can achieve multifunctional measurements: the refractive index change of the liquid medium can be detected with high sensitivity by measuring the resonant frequency of the structure. Therefore,it is possible to detect the liquid medium with such a design. Meanwhile,in terms of micro displacement sensing,a high micro displacement measurement sensitivity can be realized in both the z-axis and y-axis directions,respectively. The 3D metamaterial sensor proposed in this paper provides an idea for the design of a functionally integrated sensor in the terahertz region.

To address the problems of too many parameters and much time consumption in existing recognition networks for transmission lines,a lightweight encoder-decoder network is constructed to discern complex transmission line images featured with multiple intersections quickly and accurately. The encoder is based on the first 16 layers of conventional MobileNetV3 to reduce network parameters. The convolutional block attention module is used to replace the squeeze and excitation attention module to improve the network's ability to extract the feature information of transmission lines. The decoder is constructed by combining deeply separable convolution and deep atrous spatial pyramid pooling to expand the receptive field and improve the network's ability to aggregate contextual information with different scales. Moreover,the training network is sparse by using the L1 regularization method. The pruning threshold is determined according to the product of the scaling factor and the corresponding output of each channel to remove redundant channels and compress the network effectively,which improves the recognition speed of transmission lines. Experimental results demonstrate that the mean pixel accuracy,mean intersection over union ,and recognition speed of the lightweight encoder-decoder network are 92.11%,84.19%,and 41 frames per second,respectively,which are better than PSPNet,U2Net,and existing improved transmission lines recognition networks.

To solve the problems of low color saturation and edge blurring caused by viscous resistance and other factors in color EPD, this paper proposes a color e-paper edge enhancement error diffusion algorithm based on HSL space to improve the display quality. This algorithm first uses an edge detection operator to obtain edge-enhanced images from denoised images. It combines edge-enhanced image pixel neighborhood average gray level, pixel and neighborhood gray level difference, and pixel neighborhood similarity to obtain new RGB image pixel value. Then, the new RGB image is processed by a threshold process to obtain a 16-level RGB image. Finally, the 16-level RGB image is converted to HSL space, and a conversion model between HSL and RGB color spaces is established. According to the brightness and saturation of the pixel, the adjustment factor is calculated to enhance the saturation of the RGB image. Compared with the traditional error diffusion algorithm, the signal-to-noise ratio PSNR of this algorithm is improved by 3.9%~26.7%, the UCIQE is improved by 10.1%~48.2%, and the SSIM is improved by 13.2%~25.4%. The subjective evaluation refers to the ITU-R BT.500-1 standard to design experiments and calculate Z scores. Finally, the image details and colors of the image processed by this paper algorithm are closer to the original image on the color e-paper, and the overall visual effect is better.

Road scene semantic segmentation is a crucial task in autonomous driving environment perception. In recent years, Transformer neural networks have been applied in the field of computer vision and have shown excellent performance. Addressing issues such as low semantic segmentation accuracy in complex scene images and insufficient recognition capabilities for small objects, this paper proposes a road scene semantic segmentation algorithm based on Swin Transformer with multiscale feature fusion. The network adopts an encoder-decoder structure, where the encoder utilizes an improved Swin Transformer feature extractor for road scene image feature extraction. The decoder consists of an attention fusion module and a feature pyramid network, effectively integrating semantic features at multiple scales. Validation tests on the Cityscapes urban road scene dataset show that, compared to various existing semantic segmentation algorithms, our approach demonstrates significant improvement in segmentation accuracy.

The surface defects of solar cells exhibit significant intra-class differences, minor inter-class differences, and complex background features, making high-precision identification of surface defects a challenging task. This paper proposes a Convolutional -Vision Transformer Network (CViT-Net) that combines local and global features to address this issue. First, a Ghost-Convolution two-fusion (G-C2F) module is used to extract local features of the solar cell panel defects. Then, a coordinate attention mechanism is introduced to emphasize defect features and suppress background features. Finally, a Ghost-Vision Transformer (G-ViT) module is constructed to fuse local and global features of the solar cell panel defects. Meanwhile, CViT-Net-S and CViT-Net-L network structures are provided for low-resource and high-resource environments. Experimental results show that compared to classic lightweight networks such as MobileVit, MobileNetV3, and GhostNet, CViT-Net-S improves the classification accuracy of solar cell panels by 1.4%, 2.3%, and 1.3%, respectively, and improves the mAP50 for defect detection by 2.7%, 0.3%, and 0.8% respectively. Compared to ResNet50 and RegNet, CViT-Net-L enhances the classification accuracy by 0.72% and 0.7%, respectively, and improves the mAP50 for defect detection by 3.9% and 1.3%, respectively. Compared to advanced YOLOV6, YOLOV7, and YOLOV8 detection networks, CViT-Net-S and CViT-Net-L structures, as backbone networks, still maintain good detection performance in terms of mAP and mAP50 metrics, demonstrating the application value of the proposed algorithm in the field of solar cell panel surface defect detection.

The super-resolution reconstruction algorithm is an algorithm that restores low-resolution images to high-resolution images, which is widely applied in the fields of medicine, remote sensing, military security, and face recognition. It is hard to construct datasets in some specific scenarios, such that the application of super-resolution reconstruction algorithms based on deep learning is limited. The scanning pattern of micro-scanning imaging technology is fixed, which requires high precision of the device. To address these two problems, we propose an image super-resolution reconstruction algorithm based on active displacement imaging. Specifically, we control the camera to move randomly while recording the displacement at the sampling moment and then reconstruct the high-resolution images by solving, mapping, and selecting zones, obtaining the sub-pixel information between multiple frames, and finally iteratively updating the reconstruction. The experimental results show that this algorithm outperforms the latest multi-featured super-resolution reconstruction algorithms for POCS images in terms of PSNR, SSIM, and mean gradient. What's more, the present algorithm does not require a fixed scanning pattern, which reduces the requirement of the micro-scanning technique on the device in place accuracy.

To address the problems of blurring and insufficient sample size in sonar images, an improved sonar image target detection algorithm is proposed based on YOLOv5. The algorithm uses geometric filtering, vertical flipping, and other methods to enhance the sonar image dataset. The fusion attention mechanism module is added to make the algorithm better focus on the features of small targets in sonar images. At the same time, in response to the problem that most target detection algorithms currently run on the cloud and cannot achieve real-time sonar image detection, this paper uses lightweight network replacement and NCNN edge porting technology. It adopts the GSConv module in the neck network to successfully transplant the algorithm to the ZYNQ platform, realizing real-time detection of sonar images on the embedded end. After experiments, the algorithm proposed in this paper reduced the parameter quantity by 56%, increasing map50 and map50-95 by 2.2% and 2.5%, respectively. The algorithm’s performance has significantly improved, proving that the method proposed has certain feasibility and effectiveness in lightweight sonar image target detection tasks.

This paper first introduces the important measures of the NSFC on the reform of science fund in 2023, and then statistically analyzes the project applications and fundings of the F05 "Optics and Optoelectronics" subject, including the free category programs (general program, youth scientists fund, regional science fund), the key program, the excellent young scientists, the outstanding young scientists and other programs. It also makes a statistical analysis of the distribution of the secondary codes, supporting organizations and attributes of the four types of scientific problems of the proposals, and then introduces the pilot work situation of the review mechanism reforming of "responsibility, credibility, contribution" (RCC). Finally, combined with the project applications and funding results of this year, suggestions are given to the project applicants and the experts in the field of "Optics and optoelectronics".

Aiming at the problems of unbalanced sample distribution and difficulty identification of the lesion area in diabetic retinopathy, we propose a retinal lesions grading algorithm that integrates coordinate perception and hybrid extraction. This algorithm first processes the retinal input image and the Gaussian filtering to enhance the difference between the image lesions and the background of the noise, and then the hybrid dual models composed of the backbone network of Res2Net-50 and Densenet-121 will be enhanced. The image is extracted layer by layer to achieve the full capture of the multi-scale feature texture, then the multi-layer coordinate perception module and the attention characteristics fusion module are integrated at the mixed dual model connection to achieve the purpose of eliminating the characteristics of the lesions and the realization of different lesions. The weight of semantics is reshaped, finally uses the combined loss function to relieve the uneven distribution of samples to further supervise the training and test of the model. This article is experimented on the IDRID and Aptos 2019 data sets, with the secondary weighted coefficients of 88.76% and 90.29%, respectively. Accuracy rates were 81.55% and 84.42%, which provides a new window for the diagnosis of retinopathy grades and intelligent auxiliary diagnosis.

In order to meet the market demand of the optical industry, the rapid manufacturing technology of aspheric optical elements is one of the best solutions to realize low cost, mass production, and high precision production. This paper mainly introduces the rapid manufacturing technology of aspheric optical elements, including precision optical glass molding technology and precision optical plastic injection molding technology, and compares the two technologies with the manufacturing technology of other aspheric optical elements. This paper also expounds on the types of mold materials, mold processing methods, molding materials, molding process, etc. Finally, the development status of rapid manufacturing technology of aspheric optical elements in recent years is summarized, and future development has prospected.

Optical analog computing avoids the photoelectric conversion in various application scenarios by directly modulating the optical input in the spatial domain. Therefore, it has become a research focus in many applications such as image processing. In this paper, a polarization-multiplexed optical analog computing metasurface structure based on the Green's function method is designed using topologal optimization. Under different linearly polarized light incidence, this topological metasurface can independently tailor the amplitude and phase of the transmitted light field. It achieves bright-field imaging and one-dimensional second-order differentiation operations in orthogonal polarization states, as well as a polarization- controlled differentiation direction for a multiplexed differential system. These polarization-multiplexed designs can play a vital role in more optical computing application scenarios.

Electromagnetic metasurfaces, as a class of planar electromagnetic materials consisting of single-layer or multilayer subwavelength artificial micro-structures, can precisely control the amplitude, phase, wavefront, dispersion, polarization, and angular momentum of electromagnetic waves in the subwavelength scale. In particular, reflectionless electromagnetic metasurfaces provide new theories and schemes for realizing high-efficiency electromagnetic devices. In this review, we elaborate the underlying mechanism of reflectionless metasurfaces from the perspective of the Huygens principle, electromagnetic resonances and the Brewster effect. We also discuss the important applications including anomalous refraction, polarization manipulation, meta-antireflection coatings, and perfect electromagnetic absorption, and point out the challenges and potentials of this field.

Metasurfaces play an important role in controlling the amplitude, phase, polarization, and complex wavefront of electromagnetic waves. Dynamic tunable devices can be realized by combining various active modulation means. However, most of the existing reconfigurable devices have volatile properties that require a constant stimulus to maintain. The chalcogenide phase-change material Ge2Sb2Te5 (GST) has the characteristics of non-volatility, reconfigurability, and large optical contrast, which can be used to achieve tunable metasurface devices. In this review, we review the recent research progress of GST-based terahertz (THz) metasurface devices and introduce the spectral characteristics and reversible phase transition conditions of GST in the THz band. Furthermore, we systematically summarizes the relevant works on non-volatile, reconfigurable, and multi-level manipulation of THz amplitude, polarization, and wavefront by combining GST with metasurfaces. Finally, the future development prospects and challenges are discussed. The non-volatile nature of GST provides a new path to achieve non-volatile reconfigurable THz devices with low energy consumption, while its ultra-fast volatility can be used for next-generation high-speed communication.

With the high-speed development of mobile communication and the increasingly complex communication environment, multibeam antennas are widely required in the application fields of multi-target radar, satellite communication, multi-point wireless communication, etc. Orbital angular momentum is one of the fundamental properties of electromagnetic waves. It has a spiral wavefront and is independent of basic properties such as amplitude, phase, and polarization. It can provide a new multiplexing dimension for electromagnetic waves. Based on the excellent electromagnetic control capability of the sub-wavelength metal waveguide array, we designed a multibeam rotatable terahertz (THz) array antenna. By adjusting the phase distribution of two orthogonal polarization components of the incident wave, they can be transformed into two vortex beams with the same intensity distributions and opposite orders. The interferometric patterns in the 45° polarization direction can be rotated by changing the phase difference between the two components. Moreover, the array antenna also shows the performance of high gain (31 dBi) and wide bandwidth (up to 61 GHz). This work can provide a new way for azimuth measurement based on multibeam array antennas and is of great significance to enrich the design of array antennas in the THz band.

The quasi-bound state in the continuum (quasi-BIC) is a special resonant mode in a metasurface with a very high quality factor that can greatly enhance the light-matter interaction and has important applications in fluorescence enhancement, nanolaser, optical sensing, and nonlinear optics. In this paper, we study the application of quasi-BIC dielectric metasurface for refractive index sensing based on the theory generated by our previous quasi-BIC. The basic structure of the sensing device is given, the preparation of the sample optofluidic structure is completed by using electron beam lithography combined with injection molding process, and the performance is initially tested. The results showed that the metasurface has two high-Q quasi-BIC resonance peaks (1.523 μm and 1.570 μm, with quality factors of 3069 and 4071, respectively), thanks to the new strategy of quasi-BIC generation. The test experiments with four refractive index solutions (n=1.450/1.462/1.470/1.480, respectively) as samples showed that both resonance peaks could complete the refractive index detection with the sensitivity of 452 nm/RIU and 428 nm/RIU, respectively, and the performance evaluation indexes FOM were 376.7 and 372, respectively, which are better than the existing literature. The linearity between resonance wavelength and refractive index is good, showing the potential of quasi-BIC metasurface in refractive index sensing.

All-metal metasurfaces are structural arrays composed of sub-wavelength metal units, which exhibit high efficiency and large bandwidth in phase manipulation of electromagnetic waves. Compared with metal-dielectric hybrid metasurfaces, all-metal metasurfaces have excellent thermal and mechanical properties, such as high-temperature resistance, high strength, and good ductility, which enable them to be applied in extremely complex environments such as high temperature and high pressure. In this paper, we briefly summarize the recent research progress based on all-metal metasurfaces. We mainly introduce their applications in the construction of highly efficient and multi-functional planar optical devices as well as multi-spectrum electromagnetic stealth, and provide an outlook of the future direction of its development.

Electromagnetic metamaterials are composed of sub-wavelength artificial unit cells with periodic and aperiodic arrangements, which can achieve peculiar properties that natural materials do not have. As the two-dimensional metamaterials, metasurfaces have the advantages of low profile, easy integration and low cost. With the introduction of active elements, sensing elements and intelligent algorithms, metasurfaces further realize real-time programmable and intelligent control of electromagnetic waves. At present, most electromagnetic metasurfaces researches are devoted to the manipulation of reflecting waves and transmitting waves. In fact, electromagnetic metasurfaces also have the strong regulation ability for radiating waves. This paper will introduce the research progress of metasurfaces in regulating the amplitude, phase, polarization of radiating waves systematically. Based on the integration of metasurfaces and feeds and the regulation principle of metasurfaces on radiating electromagnetic waves, this paper focuses on folded array metasurfaces, Fabry-Perot metasurfaces, leaky wave metasurfaces and radiation-type metasurfaces, corresponding to space wave feeding, surface wave feeding, gap coupling feeding, coaxial feeding. The regulation mechanism and applications of these four types of metasurfaces on radiating waves are introduced from the perspectives of passive and active. Finally, the future research directions of electromagnetic metasurfaces in regulating radiating waves are prospected.

Surface wave (SW), as a kind of information or energy transportation platform, can find important applications in on-chip optical devices and systems. However, the efficient and free control from near-field SW to far-field propagation wave still suffers from fundamental challenges in the field of on-chip photonics. This paper starts with an introduction of the basic principles for far-field radiation. Then it reviews the approaches to control the multiple parameters (e.g., phase, amplitude, and polarization state) of the SW’s radiation field based on the metasurface, as well as the complex far-field wavefront manipulation of the surface wave, such as the directional radiation, far-field focusing, special beam excitation, and holographic imaging. Finally, the main challenges and future developments of far-field radiation control of SWs are summarized.

Metasurface can precisely modulate the fundamental properties such as polarization, amplitude, frequency, and phase of optical waves at the subwavelength scale. Based on this background, we propose and experimentally verify a multifunctional metasurface image display technology enabled by merging spatial frequency multiplexing and near- and far-field multiplexing. In near- and far-field multiplexing, the orientation degeneracy of nanostructures is introduced to combine geometric phase modulation and light intensity modulation, which leads to independent coding of near-field grayscale image and far-field holographic image displays by using simulated annealing algorithm. In spatial frequency multiplexing, different spatial frequency components of two images are added together to generate a hybrid image for hologram design. Since people receive different spatial frequency parts when the observation position changes, both high-frequency and low-frequency images can be easily distinguished. In our experiment, three independent images (a grayscale image, a high-frequency image and a low-frequency image) can be displayed simultaneously at different distances, which explains that our multifunctional metasurface has enhanced information storage capacity. This work provides a new path for multifunctional metasurface design, and possesses broad applications in optical encryption, optical anti-counterfeiting, and many other related fields.

Metasurfaces can manipulate the physical parameters of electromagnetic waves, including polarization, amplitude, and phase. The development of micro-nanofabrication technology further promotes the application prospects of metasurfaces in fields such as display, imaging, sensing, anti-counterfeiting, and optical modulation. However, most metasurfaces lack dynamic modulation, which restricts their scope of application. In recent years, the research on dynamic metasurfaces has made some progress. This review mainly introduces several mechanisms for dynamic metasurfaces, including electrical, thermal, optical, mechanical, and chemical modulations, and summarizes the research progress in the dynamic metasurfaces. In addition, this review also outlines the applications of dynamic metasurfaces in fields such as imaging, display, and optical modulations, and highlights their significance and prospects. Finally, this review summarizes the main problems and future development directions of currently tunable metasurfaces.

As a technology that combines spectral information with spatial information, spectral imaging has been widely concerned in scientific research and engineering applications. The optical field can be modulated efficiently by designing and optimizing the metasurfaces with subwavelength scale features. This article reviews the research progress of spectral imaging based on metasurfaces in recent years. Compared with traditional spectrometers, the compact spectrometers based on metasurfaces have the advantages of smaller volume and simpler optical path, and have greater application potential in small devices. According to different imaging mechanisms, spectral imaging based on metasurfaces can be divided into superdispersion, narrowband filter and broadband filter. The research progress of each imaging mechanism is introduced in detail, and then the application in practical scenarios is summarized. Finally, the development direction and application prospect of spectral imaging of metasurfaces are prospected.

Aiming at the problems of narrow working frequency band and low near field imaging efficiency in metasurface holographic imaging technology, this paper proposed the principle and model of optimization of achromatic broadband metasurface hologram imaging. A deep learning network model based on the depth image prior (DIP) is proposed for single-target passive metasurface hologram design, and achromatic broadband metasurface hologram imaging is achieved. Numerical simulation and experimental results have proved that the designed holographic imaging device can achieve good achromatic imaging effect in the 9 GHz~11 GHz frequency band, and has great potential application in the field of holographic imaging and broadband functional device design.

As an artificial micro-nano device, metasurfaces can precisely manipulate the propagation and phase of light beams. Vortex beams with different polarization vector properties possess unique optical field distribution characteristics, and the use of metasurfaces to generate complex vector vortex fields has increasingly broad research prospects. This article classifies materials for producing vector vortex beams with metasurfaces and introduces the research progress of metal metasurfaces, all-dielectric metasurfaces, and intelligent metasurfaces in vector vortex beam generation and control. We elaborate on the principles of modulating incident wavefronts using different phase theories and the characteristics of different vector vortex beams generated by metasurfaces, and explore the relationship between the two. Additionally, we summarize the advantages of using metasurfaces instead of traditional optical devices to generate vector vortex beams, and look forward to the challenges and possibilities of vector field control research using metasurfaces with different materials in the future.

The spherical aberration is an important factor affecting the resolution power of super-oscillatory telescopic systems. The reason is that the spherical aberration leads to a high sidelobe in the field of view of the intensity point spread function, which reduces the resolution of the system. In this paper, the effect of the spherical aberration on imaging in a super-oscillatory telescopic system is analyzed and the allowable range of the primary spherical aberration is determined. Based on the principle of optical super-oscillatory and the optimization method of linear programming, a super-oscillatory telescopic system is designed. A resolution of 0.68 times the Rayleigh criterion can be achieved under a working wavelength of 532 nm. A mathematical model for quantitative analysis of the super-oscillatory telescopic system with the spherical aberration is established. The system can distinguish the three-slit target under the interference of the primary spherical aberration with a root mean square (RMS) no more than 0.041 times wavelength. The imaging effect of the narrow band working wavelength in the spherical aberration system is analyzed. This paper has potential applications in optical measurement, environmental monitoring, super-resolution telescope, and other fields.

A versatile metasurface platform based on phase change materials (PCMs) is provided to realize dynamic switching between edge detection and imaging without the assistance of a 4f imaging system. The metasurface consists of a periodic arrangement of unit structures which consists of a Ge2Sb2Se4Te1(GSST) nanofin on a silicon substrate. The dynamically switchable performance results from the combination of the geometric phase and two independent propagation phases that are provided by the composed phase-change material meta-atoms in amorphous and crystalline states. The average cross-polarized transmission coefficients are 0.77 and 0.42 in the amorphous and crystalline states. In order to verify the feasibility of the design, simulation and theoretical calculation of the designed switchable metasurface are carried out in crystalline state and amorphous state respectively, which show excellent imaging and edge detection results. The proposed metasurface and its working principle have potential applications in biomedical imaging and defect detection.ect detection.

Infrared-visible person re-identification has been widely used in video surveillance, intelligent transportation, security, and other fields. However, due to the differences between different image modalities, it brings great challenges to this field. The existing methods mainly focus on mitigating the differences between modes to obtain more discriminating features, but ignore the relationship between adjacent features and the influence of multi-scale information on global features. Here, a infrared-visible person re-identification method (MFANet) based on multi-feature aggregation is proposed to solve the shortcomings of existing methods. Firstly, the adjacent level features are fused in the feature extraction stage, and the integration of low-level feature information is guided to strengthen the high-level features and make the features more robust. Then, the multi-scale features of different receptive fields of view are aggregated to obtain rich contextual information. Finally, multi-scale features are used as a guide to strengthen the features to obtain more discriminating features. Experimental results on SYSU-MM01 and RegDB datasets show the effectiveness of the proposed method, and the average accuracy of SYSU-MM01 dataset reaches 71.77% in the most difficult all-search single-shot mode.

Laser alignment is a prerequisite for stable energy acquisition at the receiver end in laser wireless power transmission systems. Laser power transfer imposes high requirements on alignment accuracy, stability, and real-time performance. Therefore, a laser alignment system design method is proposed, and optimizations are made to the region of interest extraction and image preprocessing methods. On one hand, the SSD (single shot multi-Box detector) model is improved by introducing MobileNet, incorporating spatial attention mechanism, and fusing semantics. The improved model is used for training and predicting the regions of interest. Compared to the original model, the training speed is improved by 71.67%, the model size is reduced by 52.48%, the real-time detection speed is increased by 295.30%, and the detection error is significantly reduced. On the other hand, the weights of grayscale conversion are optimized, and an adaptive threshold selection using a histogram is implemented. The elliptical fitting method and centroid method are employed to detect the spot and beacon center, reducing the error in spot localization. Experimental results show that the improved laser alignment system achieves a stable accuracy of over 95% and meets the requirements of accuracy, speed, and stability in application.

A 2.8 μm passively Q-switched mode-locked erbium-doped fluoride fiber laser based on material saturable absorption is reported in this paper. By depositing TiCN particles directly onto the cavity mirror as the saturable absorber and using the vertical cleaved end of the fluoride fiber as an output coupler, the 2.8 μm pulsed fiber lasing with a low laser threshold and a compact cavity structure is realized. When the pump power reaches 330 mW, the Q-switched mode-locked pulses begin to appear. With the increase of pump power, the repetition frequency of Q-switched pulse envelope increases from 14.34 to 32.57 kHz, and the pulse width decreases from 10.51 to 5.40 μs. Under the pump power of 650 mW, the maximum average output power of 25.83 mW is obtained, and the slope efficiency is about 7.2%.

Over the past few years, the field of micro-/nano- photonics has witnessed a surge in research focused on developing innovative optical devices that offer dynamic spectra engineering. Among the materials showing promise in this area, vanadium dioxide (VO2) can actively manipulate its refractive index via a phase transition process, enabling the dynamic manipulation of spectra. In this work, a photosensitive polymer nanocomposite with tunable effective refractive index is prepared by incorporating VO2 nanocrystals into methacrylate monomers, which takes advantages of the phase change characteristics of VO2 and the photopolymerization properties of the monomer. In addition, with the aid of the state-of-the-art femtosecond laser processing technology, highly precise two-dimensional and three-dimensional micro-/nano- optical structures embedded with the phase change capabilities outlined by VO2 are achieved. Fascinatingly, the spectra measurements via Fourier transform infrared spectrometer reveal that when subjected to the critical phase transition temperatures, the printed micro-/nano- structures will undergo a thermally induced phase transition of the VO2 nanocrystals embedded within them. Consequently, there is a discernible alteration in the effective refractive index of the optically functionalized structure, inspiring the dynamic manipulation of the short-band spectra.

Optical multilayer interference tomography (OMLIT) is used in correlated light and electron microscopy to image a large field through an optical microscope, providing the navigation of the region of interest for later nanometer-resolution electron microscope imaging. In order to further improve the imaging contrast and positioning accuracy of thin film samples, a theoretical model combining polarization illumination and OMLIT is proposed. This model is written in the matrix formalism and the propagation of polarized light through different layers with various incident angles is simulated. The simulation results show that using the polarized light with an electric field oscillating parallel to the incidence plane (p-polarization) exhibits a much higher imaging contrast than the unpolarized light. Especially when the p-polarized light illuminates on an OMLIT sample, of which the first coating layer is Ag, with an incidence angle of 62°, the imaging contrast can be vastly enhanced by 138 times. The presented model provides a theoretical basis for polarization illumination OMLIT, pathing a new technical way for the development of the correlated light and electron microscopy technique.

The silicon-based on-chip power splitter, an important component of the photonic integrated circuits, has much wider scope of applications such as feedback circuits, tap-port power monitoring, and optical quantization. The design methods of the nanophotonic devices can be roughly divided into the forward design and inverse design methods. This review article outlines the differences and connections between the forward design and inverse design methods, and classifies the inverse design algorithms. In addition, the review article summarizes the representative inverse-designed silicon-based on-chip power splitters in recent years, including multichannel power splitters, arbitrary-split-ratio power splitters, multimode power splitters, broadband power splitters, and multifunction power splitters. Finally, the summary and outlook are made on the development trend of the inverse design algorithms and the inverse-designed power splitters.

Lane detection is a challenging task due to the diversity of lane lines and the complexity of traffic scenes. The detection results of the existing detection methods are not ideal when the vehicle is driving in congestion, at night, and the lane lines are not clear or blocked on the road such as curves. Based on the framework of detection methods, a method that axial attention-guided anchor classification lane detection is proposed to solve two problems. The first is the problem of missing visual cues when lane lines are unclear or missing. The second problem is the lack of feature information caused by using sparse coordinates on mixed anchors, which leads to a decline of detection accuracy. Therefore, an axial attention layer is added to the backbone network to focus on prominent features of the row and column directions to improve the accuracy. Extensive experiments are conducted on the TuSimple and CULane datasets. Experimental results show that the proposed method is robust under various conditions while showing comprehensive advantages in terms of detection accuracy and speed compared with existing advanced methods.

Aiming at the problem that vehicle models are difficult to recognize due to differences in vehicle posture and viewing angles, a vehicle model recognition network based on progressive multi-granularity ResNet is proposed. Firstly, a progressive multi-granularity local convolution module is proposed by using the ResNet network as the backbone network to perform local convolution operations on vehicle images of different granularity levels, so that local features of vehicles at different granularity levels can be paid attention to when the network is reconstructed. Secondly, for the multi-granularity local feature map, the random channel discarding module is adopted to perform random channel discarding, which suppresses the network's attention to the vehicle's salient regional features and improves the attention of non-salient features. Finally, a progressive multi-granularity training module is proposed. A classification loss is added in each training step to guide the network to extract more discriminative and diverse vehicle multi-scale features. Experimental results show that the recognition accuracy of the proposed network reaches 95.7%, 98.8%, and 97.4% respectively on the Stanford-cars dataset, the Compcars network dataset, and the vehicle model dataset VMRURS in real scenes. In comparison with the comparative network, the proposed network not only has higher recognition accuracy but also has better robustness.

Off-axis reflector telescopes are mainly used in space astronomy observation and other related fields. The image quality of off-axis two-inversion telescopes is sensitive to lens misalignment and is more difficult to calibrate using a laser interferometer after misalignment in an operating environment. To address this challenge, this paper proposes a method that uses the out-of-focus spot map of the system for infinity point targets and uses the Swin-Transformer network to calculate the amount of lateral misalignment of the secondary mirror. Theoretical calculations were used to determine the camera defocus positions that could avoid the multi-solution problem, and simulations were used to investigate the effect of different defocus amounts on calibration accuracy. The trained network uses a frame of out-of-focus spot map of the out-of-tune system to perform the estimation of the amount of out-of-tune. Both simulation analysis and experimental results verify the effectiveness of the method, which can achieve high accuracy and fast correction of out-of-tune telescope systems in the working environment.

Optical skyrmions provide a new idea and approach to endow structured light and spatial-temporal light with topological properties. In this paper, the longitudinal and transversal components of the focused light field are decoupled and can be controlled independently by modulating the polarizations and phases of two pairs of counter-propagating incident cylindrical vector beams under 4π focal configuration. Under this condition, Néel-type and Bloch-type optical skyrmions formed by electromagnetic field vectors are constructed in the focal plane. When one pair of the incident beams is radially polarized with a phase difference of π and the other pair is radially polarized in phase, a Néel-type optical skyrmion formed by electric field vectors can be constructed in the focal plane of the 4π focal system. The corresponding focal magnetic field is purely azimuthally polarized. If we substitute the other pair of the incident beams with azimuthally polarized beams, Bloch-type optical skyrmions formed by electromagnetic field vectors can be constructed in the focal plane, where the one formed by magnetic field vectors has a π/2 phase lead compared with the that formed by electric field vectors. This work provides a theoretical basis for further research on the interactions between matter and optical skyrmions formed by electromagnetic field vectors at micro and nano scales in free space.

Surface enhanced Raman spectroscopy (SERS) is a kind of molecular spectrum, which has the characteristics of rapidity, high sensitivity and fingerprint recognition, and has important applications in analytical chemistry, biomedicine and other fields. However, some detection molecules in the solution sample are difficult to be adsorbed by the SERS substrate, resulting in difficulty in enhancing the Raman signal of molecular. To this end, this paper proposes a core-shell structure (AuNRs@ZIF-8) of ZIF-8 material coated gold nanorods (AuNRs) to achieve Raman signal enhancement. Both the surface plasmon enhancement characteristics of the gold nanoparticles and the adsorption properties of ZIF-8, a porous MOFs material, can be used to realize highly sensitive Raman detection of the solution samples. We first prepare the well-homogeneous AuNRs by seed crystallization method, then modify them with polyvinylpyrrolidone (PVP), and finally add the metal-organic framework ZIF-8 precursor to obtain the AuNRs@ZIF-8 Core-shell nanostructures. The structure has high sensitivity to the SERS detection of rhodamine (R6G), the detection limit can be as low as 10?9 M, and the linear relationship and homogeneity are good. In addition, we further confirm the formation of the core-shell nanostructures and effective adsorption of the target molecules by testing the UV-Vis absorption spectra of the structure before and after R6G absorption.

As a kind of underwater active sonar equipment, forward-looking sonar is often used to collect underwater image data. However, it will be affected by underwater noise, which leads to the degradation of image quality. Due to this situation, a forward-looking sonar image denoising method is proposed based on dense residuals and a dual-channel attention mechanism network. Firstly, the two-channel attention mechanism is adopted to extract the channel information of the sonar image, collect the global information of the sonar image, and output the noise map of the sonar image. Based on the noise image and sonar image, the dense residual block fully learns the feature information at different scales and outputs a clean sonar image after multiple learning and information transfer. Because of the forward-looking sonar image and its noise characteristics, the forward-looking sonar image is simulated and the multiplicative noise of Rayleigh distribution and the additive noise of Gaussian distribution are added to generate a simulation dataset for network training and performance evaluation. Experimental results on the simulated data set and real data set show that the proposed method can effectively remove the noise and retain image details.

Multiple object tracking (MOT) is an important task in computer vision. Most of the MOT methods improve object detection and data association, usually ignoring the correlation between different frames. They don’t make good use of the temporal information in the video, which makes the tracking performance significantly degraded in motion blur, occlusion, and small target scenes. In order to solve these problems, this paper proposes a multiple object tracking method with the aligned spatial-temporal feature. First, the convolutional gated recurrent unit (ConvGRU) is introduced to encode the spatial-temporal information of the object in the video; By considering the whole history frame sequence, this structure effectively extracts the spatial-temporal information to enhance the feature representation. Then, the feature alignment module is designed to ensure the time consistency between the historical frame information and the current frame information to reduce the false detection rate. Finally, this paper tests on MOT17 and MOT20 datasets, and multiple object tracking accuracy (MOTA) values are 74.2 and 67.4, respectively, which is increased by 0.5 and 5.6 compared with the baseline FairMOT method. Our identification F1 score (IDF1) values are 73.9 and 70.6, respectively, which are increased by 1.6 and 3.3 compared with the baseline FairMOT method. In addition, the qualitative and quantitative experimental results show that the overall tracking performance of this method is better than that of most of the current advanced methods.

The Solar-blind UV detection has wide application scenarios and unique market values in the civil and military fields, such as space security communication, ozone hole detection, missile attack warning and so on. Gallium oxide (Ga2O3) has an extremely wide band gap (4.4-5.3 eV), almost covering the entire solar-blind UV region, and is considered as one of the most promising materials for the preparation of solar-blind UV photodetectors. Compared with single crystal or epitaxy materials, amorphous gallium oxide (a-Ga2O3) has a lower deposition temperature, a relatively simple preparation process, and a much wider range of applicable substrates. Therefore, it has become a new research hot spot in the field of the Ga2O3 solar-blind UV detection in most recent years. In this paper, the basic characteristics and most common preparation methods of a-Ga2O3 are introduced firstly, and then the research progress and present situations of the a-Ga2O3-based solar-blind UV photodetector are introduced in details from the perspective of device structures. At present, a-Ga2O3 based solar-blind UV photodetectors are mainly divided into MSM, junction, TFT and array types. By the optimization of device structures, the photodetection performance has been significantly improved. MSM device is the most widely used because of its simple structure and high responsivity. By constructing Schottky junction or heterojunction, the junction-type devices own the characteristics of fast response speed, low dark current, and self-power supply. TFT devices can suppress the dark current, amplify the gain and improve the recovery speed by applying gate voltage. Array-type devices can be used for large-area imaging. Finally, the future development trends of the a-Ga2O3 solar-blind UV photodetector are summarized.

Three-dimensional point cloud reconstruction of underwater targets has important applications in underwater exploration and obstacle avoidance of underwater unmanned vehicles. In this paper, the underwater target detection experiment is carried out by simulating the temperature and salinity water turbulence in the laboratory, and the influence of temperature difference and salinity difference on the laser point cloud three-dimensional reconstruction effect of underwater submarines, gliders and anchor mines are studied, and the error analysis of the reconstructed 3D point cloud data is carried out. The results show that in the process of underwater target 3D point cloud reconstruction, with the enhancement of water temperature turbulence and salinity turbulence intensity, the effective points of the reconstructed point cloud data decrease significantly, and the mean error increased significantly. The research results have a certain reference value for the development of underwater three-dimensional point cloud reconstruction system.

A side-scatter lidar is known to have evident advantages over other types of lidar in atmosphere detection. However, the signal of the side-scatter lidar may suffer from the noise as all other lidars. It is noted that the original signal of the side-scatter lidar is an image captured by a CCD camera. Therefore, denoising the side-scatter lidar signal may need more efforts than ordinary radar signals. In the paper, a denoising algorithm based on convolution neutral network is proposed for the side-scatter lidar signal. We combine the residual learning with batch standardization in the network. Further, attention mechanism and activation function in the network are optimized in order to improve the learning efficiency and the network output performance. Using the proposed algorithm, we successfully identify the noise and separate the noise from the simulated lidar signal. The signal-to-noise ratio is hence increased. Simulation results show that the peak signal-to-noise ratio is increased by over 5 dB using the proposed denoising algorithm. The relative error of signal is reduced to 9.62%. The proposed denoising algorithm based on the convolution neutral network is shown to be efficient for improving the side-scatter lidar signal, compared with the possible denoising algorithms based on wavelet transform and Wiener filtering.

The frequency-modulated continuous wave laser interferometry is widely used in the field of precision measurement. Aiming at its high-precision displacement demodulation problem, this paper applies the centroid method to the demodulation of its beat signal, proposes a phase demodulation algorithm based on the centroid peak-seeking method, and carries out experiments and analysis. Based on the smooth filtering and peak clipping of the intercepted beat signal, the proposed algorithm obtains the centroid of the beat signal through the centroid coordinate formula. The abscissa of the centroid is the peak position. Finally, the phase discrimination algorithm demodulates the displacement. In the simulation, the SNR is set to 15 dB, the phase error of the algorithm is 0.016 rad, and the displacement error is 2.04 nm. A frequency-modulated continuous wave laser interference displacement measurement system was built for experimental verification. The experimental results show that when the fixed distance is 44 mm, the standard deviation of random displacement error is 2.18 nm. Compared with the conventional zero crossing detection method, the measurement error of the algorithm is reduced by 49%, the resolution is improved, and the algorithm has broad application prospects.

Phase diversity is one of the commonly used image post-reconstruction methods. In order to improve the robustness of phase diversity for solar image reconstruction, this paper proposes an improved phase diversity method based on the low-rank prior, i.e., the nuclear norm regularization of the image is introduced into the phase diversity model, and the image sub-model and the phase sub-model are solved by the half-quadratic splitting method and BFGS respectively. Reconstruction experiments and analysis are carried out on the simulated degraded focused and defocused solar images. Compared with the classical phase diversity based on Tikhonov regularization, the phase diversity based on low-rank prior can improve the accuracy of wavefront phase estimation and the quality of reconstructed images in terms of subjective visual effects and objective indexes in both the noise-free and noise-included cases.

To solve the problem of beam splitting limitation of cemented cubic beam splitters in the mid-wave infrared band, the optical design scheme of mid-wave infrared Fizeau interferometers based on wedge splitting is proposed. At the working wavelength of 3.39 μm, to reduce the return error of the interference system and improve the measurement accuracy, a two-reflection folding collimating optical path structure is adopted, which not only ensures a good collimating wavefront, but also optimizes the design of the optical wedge to take into account the wavefront quality of the interference imaging. ZnSe and CaF2 materials are used, the collimator of the interferometer is a single plano-convex aspheric structure, and the imaging lens is composed of two separate spherical mirrors. Through the Montecarlo simulation tolerance analysis, the collimator wavefront PV of the collimator within 0.1° field of view is better than λ/4. The normalized field of view imaging wavefront PV of the interferometric optical path is better thanλ/5; The interferometric system return error is smaller than λ/50 at 0° field of view placed on the standard surface and the surface under test is tilted within 0.05°.

Polarization integrated detector has the advantages of small size, light weight, compact structure, and no need for image registration to detect and recognize the dynamic targets simultaneously at the same place. This article mainly introduces the research progress and applications of polarization integrated infrared detector. We analyse the key technologies for obtaining high extinction ratio polarization integrated detectors, such as grating structure design and simulation, submicron polarization grating preparation, integration and testing, polarization image data reconstruction, etc. Finally, we introduce the typical applications of polarization imaging in unmanned aerial vehicles, camouflaged trucks, landmines, sea vessels, facial recognition, road recognition, sea oil leakage detection, medical detection, etc.

Non-line-of-sight imaging techniques can be used to capture images of objects that are hidden away in corners, and this technology has many useful applications. The captured signal is insufficient when the laser sends laser light to the intermediate surface, which is a problem when utilizing a digital micromirror device (DMD) in a confocal non-line-of-sight optical route because of the spectroscopic impact of the DMD. In order to increase the signal-to-noise ratio of the echo signal, the acquisition method of spatial multiplexing detection (SMD) is introduced into the confocal optical path in this study. SMD is integrated with the reconstruction algorithm of the light-cone transform (LCT). Based on the confocal non-line-of-sight imaging optical path, the laser initially emits laser pulses, followed by SMD collecting concealed object echoes, and the LCT algorithm concludes the reconstruction. The experimental results demonstrate that the single-pixel camera can complete the reconstruction using the LCT algorithm after employing the SMD acquisition method in the confocal non-line-of-sight optical route, which may improve the signal-to-noise ratio and the reconstruction quality.

Scintillation glasses have potential application in hadron calorimeter of circular electron positron collider (CEPC) due to the advantages of simple preparation process, flexible and controllable size, and low cost. Among them, Ce3+ luminescent center doped scintillation glasses have better scintillation performance. Matrix glass can be classified into oxide glass, halide glass, and glass-ceramic. According to the classification of different matrix glasses, this paper focuses on the optical transmittance, light yield, decay time, and radiation resistance properties of the Ce3+-doped scintillation glasses. Moreover, we introduce and summarize the research progress at domestic, foreign, and GS R&D Group. In view of the research status of different glasses, the methods for improving the glass performance are discussed from two aspects of glass composition and preparation. Finally, the future research and development directions of Ce3+-doped scintillation glasses are prospected.

The lightweight SBG inertial navigation system has the characteristics of being small and light, which can be used for the direct and strapdown stabilization of the electro-optical tracking system. In this paper, five inertial stabilization control methods based on the lightweight SBG navigation system in gimbals are studied, and theoretical analysis and experimental validation are performed. The classical position strapdown stabilization technique has limited disturbance suppression capability due to the limitation of position bandwidth, which makes it difficult to meet the requirement of high-precision stabilization. A SBG-based position-rate dual strapdown feedforward stabilization method is proposed, and the decoupling of the strapdown feedforward is achieved by introducing a high-pass filter, which can further improve the perturbation rejection bandwidth as well as the stabilization capability of the system. Theoretical analysis and experimental results show that the SBG-based gyro direct stabilization control method is better at disturbance suppression when the application conditions are not considered, although it is limited by the bandwidth. Under the limitations of platform volume and weight, the proposed SBG-based dual disturbance strapdown stabilization can further improve the system's disturbance rejection capability and obtain better stabilization accuracy.

With the rapid development of military optoelectronic technology, the role of stealth technology in modern combat systems is becoming more and more important, among which, stealth materials are crucial to improve stealth performance. We focus on infrared stealth materials and review the research progress of domestic and foreign infrared stealth materials from three aspects including single-band infrared stealth, multi-band compatible infrared stealth, and dynamic infrared stealth, and provide an in-depth analysis on the large-area flexible processing methods for micro-nano structures. The main problems of current infrared stealth materials are summarized and the future development trend is foreseen. In the future, to achieve high-performance stealth functions, new infrared stealth materials will further develop in the direction of high strength, large area, flexibility, and intelligence.

In a traditional epi-illumination fluorescence microscope, the detection path is coaxial with the illumination path, which induces the non-focal plane fluorescence and deteriorates the imaging quality. Light sheet fluorescence microscopy (LSFM), differing from the traditional fluorescence microscopes, adopts an orthogonal configuration of detection and illumination paths. A thin sheet is formed from the excitation beam, which only excites a single layer of the sample. This methodology prevents the excited fluorescence from the non-focal plane during imaging. Besides, the utilization of the laminar illumination light can significantly reduce the exposure time of the fluorescence imaging. As a result, the effects of photobleaching and phototoxicity are decreased. In this review, we first introduce the basic light path structure compositions of a LSFM system as well as the optimization and innovation based on these structures. Next, we discuss enormous processing methods developed for samples both in vitro and in vivo. Benefiting from all these innovations, LSFM outstands in performing the 3D imaging of the fluorescence-labeled biological samples and can function steadily for a long recording time. Finally, we propose potential researching directions in the future, and discuss the technical limitations of current LSFM. This review aims to provide researchers in the relevant scientific research fields with a comprehensive understanding and inspiring reference of LSFM techniques.

The lithography objective is the core component of the lithography machine, and its wave aberration determines the resolution and overlay accuracy of the lithography machine. With the gradual improvement of the performance of the lithography machine, the wave aberration requirement of the lithography objective lens has been reduced to below 0.5 nm (RMS), which is a great challenge to the detection of the wave aberration. The detection accuracy of current lithography objective wave aberration detection methods (such as Hartmann method, shear interference method and point diffraction method, etc.) is often limited by its systematic error, and absolute detection technology is a method that can separate the systematic error. The technology that came out finally broke the limit of precision. This paper reviews the wave aberration detection method and surface absolute detection technology of lithography objective lens system, combs the application and research progress of absolute detection technology in wave aberration detection in detail, and summarizes the application of absolute detection technology in different wave aberration detection methods. At the same time, combined with these difficulties, the future development trend of the absolute detection technology of wave aberration of lithography objective lens is prospected.

In order to realize the rapid measurement of composite surfaces with diffuse and mirror reflection, the composite surface measurement system based on fringe projection and fringe reflection can obtain the absolute phase rapidly through the multi-color channel of the camera. Aiming at the crosstalk and chromatic aberration between the color channels introduced by the camera, projector, and display in the composite surface topography measurement, this paper studies the crosstalk elimination method based on the matrix and the chromatic aberration elimination method of the absolute phase corresponding pixel deviation. Based on the crosstalk matrix, the crosstalk matrix of the projector and display screen is established. The crosstalk intensity from other channels in the desired color channel is eliminated to complete the crosstalk elimination between color channels. The absolute phase in the horizontal and vertical directions of each color channel is obtained by color orthogonal stripes. The relationship between phase difference and pixel deviation is established to realize the pixel deviation correction of each pixel point and eliminate the influence of color difference. The experimental results show that the proposed method reduces the average measurement error of the composite step from 0.479 mm to 0.030 mm, and improves the efficiency and accuracy of measurement.

The voice coil motor-driven fast steering mirror is an important part of a high-precision photoelectric tracking system. In the photoelectric tracking system of the moving platform, the fast steering mirror system will suffer more complex and intense internal and external interference. The traditional passive interference suppression methods and the active interference suppression methods that treat the interference as lumped interference will not be enough to ensure the high-precision stability of boresight. Therefore, this paper proposes a sliding mode composite layered interference observation and compensation control strategy which combines harmonic interference observation and extended state observation. Firstly, the harmonic disturbance observer is used to observe the harmonic disturbance with a priori frequency information. Then the extended state observer is used to observe other unknown disturbances. Finally, based on the observed multi-source interference, the sliding mode nonlinear method with anti-interference ability is used to design a composite controller to maximize the suppression of multi-source disturbances suffered by the system. The experiment shows that the sliding mode composite layered interference observation compensation method proposed in this paper can significantly improve the LOS stability accuracy of the fast steering mirror compared with the traditional single interference observation compensation method.

As a kind of high-quality factor thermo-electromechanical coupling device composed of isotropic elastic-optical crystal and piezoelectric crystal, photoelastic modulator (PEM) is widely applied for polarization measurement, spectrum measurement, and many other purposes. However, the resonant frequency tends to drift with temperature changes in the high-voltage resonant state, which destabilizes the phase modulation amplitude of the photoelastic modulator and reduces the driving efficiency. To solve this problem, the resonant frequency characteristics of the photoelastic modulator are analyzed at first. Then, a compound resonant network model of the photoelastic modulator and its high voltage resonant driving circuit is established, and a solution to frequency tracking based on the amplitude-frequency characteristics of the resonant network is proposed. Besides, a control test system based on field programmable gate array (FPGA) is developed to achieve resonant frequency tracking and modulation amplitude measurement. The test results show that this method is applicable to track the resonant frequency effectively and improve the stability and driving efficiency of the elastic light modulator. The duration of the test exceeds 90 min, and the standard deviation of the phase modulation amplitude is 0.83% rad.

Feature extraction in the CNN-based stereo matching models has the problem that it is difficult to learn global and long-range context information. To solve this problem, an improved model STransMNet stereo matching network based on the Swin Transformer is proposed in this paper. We analyze the necessity of the aggregated local and global context information. Then the difference in matching features during the stereo matching process is discussed. The feature extraction module is improved by replacing the CNN-based algorithm with the Transformer-based Swin Transformer algorithm to enhance the model's ability to capture remote context information. The multi-scale fusion module is added in Swin Transformer to make the output features contain shallow and deep semantic information. The loss function is improved by introducing the feature differentiation loss to enhance the model's attention to details. Finally, the comparative experiments with the STTR-light model are conducted on multiple public datasets, showing that the End-Point-Error (EPE) and the matching error rate of 3 px error are significantly reduced.

Learning with limited data is a challenging field for computer visual recognition. Prototypes calculated by the metric learning method are inaccurate when samples are limited. In addition, the generalization ability of the model is poor. To improve the performance of few-shot image classification, the following measures are adopted. Firstly, to tackle the problem of limited samples, the masked autoencoder is used to enhance data. Secondly, prototypes are calculated by task-specific features, which are obtained by the multi-scale attention mechanism. The attention mechanism makes prototypes more accurate. Thirdly, the domain adaptation module is added with a margin loss function. The margin loss pushes different prototypes away from each other in the feature space. Sufficient margin space improves the generalization performance of the method. The experimental results show the proposed method achieves better performance on few-shot classification.

Aiming at the problems of color distortion, noise amplification, and loss of detailed information in the process of low illumination image enhancement, a progressive fusion of parallel hybrid attention (PFA) is proposed. First, a multi-scale weighted aggregation (MWA) network is designed to aggregate multi-scale features learned from different receptive fields, promote the global representation of local features, and strengthen the retention of original image details; Secondly, a parallel hybrid attention module (PHA) is proposed. Pixel attention and channel attention are combined in parallel to alleviate the color difference caused by the distribution lag of different branches of attention, and the information between adjacent attention is used to complement each other to effectively improve the color representation of images and reduce noise; Finally, a progressive feature fusion module (PFM) is designed to reprocess the input features of the previous stage from coarse to fine in three stages, supplement the shallow feature loss caused by the increase of network depth, and avoid the information redundancy caused by single stage feature stacking. The experimental results on LOL, DICM, MEF, and LIME datasets show that the performance of the method in this paper is better than that of the comparison methods on multiple evaluation indicators.

How to further improve the lightweight ratio of high-performance meter-level space mirrors is one of the core issues in the field of large aperture optomechanical structure design. In this paper, a primary mirror with a clear aperture of Φ1200 mm was developed for a high-resolution space camera, which achieves the goal of designing area density below 40 kg/m2. The SiC mirror blank was prepared by gel injection molding and reaction sintering process. The state of the optical axis being horizontal was taken as the testing state to simplify the supporting structure. Novel lightweight measures such as an alternate arrangement of main and auxiliary stiffeners and the addition of lightweight holes on vertical walls were used inside the semi-closed mirror blank. The distributed datums were used to replace traditional datum settings, which reduces the machining area of datums by more than 80% and improves machining efficiency. Through parametric modeling and integrated optimization, the optimal structural parameter combination of the mirror blank was determined, with the final design weight of 46.9 kg. The RMS value of self-weight deformation of the mirror blank is only 2.87 nm under the state of the optical axis being horizontal, and its free fundamental frequency is 602 Hz, indicating that the primary mirror proposed in this paper has good dynamic and static characteristics. After machining, the measured weight of the mirror blank is 51.3 kg, about 9.4% overweight, and the facesheet is about 1-mm thicker than the design. At present, the mirror has already been polished to RMS λ/8 (λ=632.8 nm) of surface shape accuracy, with no obvious print-through effect observed.

Laser shock peening uses the force effect of the laser to strengthen the surface. The traditional laser shock peening technology is a single-sided shock. When applied to thin-walled parts with complex profiles, it is difficult to achieve shape control and fatigue performance control coordination. The new double-sided laser shock peening technology is ideal for solving the surface strengthening challenges of thin-walled parts with complex profiles. On the basis of introducing the characteristics and deficiencies of single-sided laser shock peening technology, the principle and technical characteristics of two double-sided laser shock peening technologies are summarized. The application of simulation research in analyzing the physical mechanism of stress wave propagation and stress field distribution of double-sided laser shock peening is expounded. The mechanism and application of double-sided laser shock peening in the application of shape control and fatigue performance control are introduced, and the future development of double-sided laser shock peening is prospected.

Under the opportunity of "Made in China 2025", in the field of ultra-precision, China has broken through many key bottleneck technologies, achieved many remarkable scientific research results, built a number of high-level ultra-precision processing technology innovation platforms, talent growth platforms and application demonstration bases, and created an independent ultra-precision industry in China. This paper mainly introduces the research progress of optical ultra-precision processing technology and equipment in the Precision Engineering Laboratory of Xiamen University. Focusing on the grinding and polishing processing of large-diameter optical aspherical components, the processing technology, grinding and polishing equipment, equipment monitoring and control software and related unit technologies developed by the group are described. These research results can provide manufacturing and processing technology support and equipment solutions for the ultra-precision processing of high-end optical components.

Surface-enhanced Raman scattering (SERS) provides important applications in diverse fields. In order to achieve high-precision SERS detection of trace molecules, current research focuses on how to increase the density of hot spots and the number of analyte molecules in the detection area. An ultrafast laser can rapidly construct large-area micro/nano-structures on material surfaces. It is important for the commercial preparation of high-performance SERS sensors. In this paper, the ultrafast laser preparation of high-performance SERS sensors is introduced from the aspect of the density of hot spots and the number of analyte molecules in the detection region. Ultrafast lasers enable both "bottom-up" and "top-down" processing. In particular, the superhydrophobic surface prepared by the ultrafast laser is one of the most effective methods to achieve the enrichment of analyte molecules. Finally, a prospect for the development of laser-prepared SERS substrates is provided.

Due to the stable mechanical and chemical properties, excellent photoelectric properties, and other advantages, hard and brittle materials have been widely used in aerospace, the photoelectric industry, and other fields. Laser fabrication is an ideal technology for hard and brittle materials processing due to its high precision, high energy, and non-contact properties. In order to achieve the removal of hard and brittle materials, high laser energy is usually required, resulting in low structural accuracy and poor surface quality. This review introduces the advances of liquid-assisted laser fabrication technology in the processing of hard and brittle materials, introduces the principles of three liquid-assisted laser fabrication technologies, and compares their advantages and disadvantages. The effects of different processing technologies, types of auxiliary liquids, and processing parameters on the quality of different materials were summarized in detail. The main applications of liquid-assisted laser fabrication technology were summarized, and the existing problems and potential development of this technology were discussed.

Functional surface with specific wettability is one of the indispensable means for droplet manipulation. In recent years, the photo-responsive functional surface with changeable wettability has developed fast. By inducing wetting gradient force, mechanical deformation, phase transformation, dielectric electrophoresis force and electro wettability alteration on the material surface, the behavior of the droplets can be controllably manipulated by the photo-responsive functional surface. In this paper, the development of the photo-responsive functional surface for the droplet manipulation was briefly reviewed. The principles and mechanisms of the droplet manipulation with the functional surface had been expatiated. The categories, structural characteristics and corresponding preparation techniques of the functional surface were analyzed and summarized. In addition, the applications of the photo-responsive functional surface in droplet transportation, fusion, fission, liquid robot, and fluidic chips were introduced in detail. The development tendency and potential applications of the photo-responsive droplet manipulation functional surface were prospected in combination with the characteristics of the functional surface.

Surface-enhanced Raman spectroscopy (SERS) technique plays an important role in molecular recognition fields due to its highly sensitive and high-resolution. As an emerging low-cost, high machining accuracy, and high-flexibility processing method, femtosecond laser direct writing processing has been widely used in the field of preparing SERS substrates. This work introduces four methods of preparing SERS substrates by femtosecond laser direct writing, including femtosecond two-photon reduction, femtosecond laser cutting metal, femtosecond laser cutting-sputtering, and femtosecond laser 3D printing. The article introduces the performance and application scenarios of each method in preparing SERS substrates and illustrates the advantages of femtosecond laser direct writing processing in preparing SERS substrates, aiming to provide a reference for future related research.

Laser microfabrication has the characteristics of ultra fast, ultra precision, etc. It has unique advantages in the field of medical equipment by contrast with traditional processing technology. Especially, it plays an irreplaceable role in the surface processing of biological materials to improve the biocompatibility of materials. The latest application of laser microfabrication in the field of medical equipment manufacturing and processing in recent years is reviewed. The structure and surface manufacturing of vascular stents and bone stents, and the surface modification of biomaterials and antibacterial treatment are emphatically introduced. Finally, the limitations of the current laser micromachining technology are discussed, and the application and development of laser micromachining technology in medical equipment in the future are prospected.

In this paper, we propose a femtosecond laser-induced surface periodic structure (LIPSS) based on relaxed ferroelectric PMN-PT crystals. By changing the applied laser parameters, the period of the LIPSS structure is varied from 750 nm to 3 μm. Finally, the phase transition properties of the LIPSS structure are investigated by increasing the temperature. Compared to the phase transition properties of the substrate, the Curie temperature of the LIPSS structure is significantly reduced, and this will provide a possible new idea for the preparation of temperature-controlled modulators based on PMN-PT crystals.

Two-photon polymerization (TPP) based on femtosecond laser has been a research hotspot in 3D micro/nano writing technology. With the increasing demand for processing complex and large-scale miniaturized 3D devices in the fields of life science, material engineering, micro and nano optics, and etc., the issue of low processing efficiency of TPP is becoming increasingly serious. During the long fabrication period, many disturbances can be introduced in the processing, causing the quality deterioration of the structure and seriously hindering the further popularization and application of these crucial 3D devices. This paper respectively compares the four approaches of single-beam writing, parallel multi-beam writing, pattern projection, and 3D projection exposure based on the TPP lithography efficiency. Moreover, the researches on the optical design of system, the writing accuracy, the fabrication throughput, the writing strategy, and etc. of each approach are also described. And the advantages and disadvantages of these four methods are summerized simultaneously. Finally, we also made a brief prospect to the developing trend of TPL efficiency improvement in the future.

Miniaturization, integration, and flexible deformation are the future development trends of optical devices. Meanwhile, optical systems based on integrated micro-optical devices stand out for their low power consumption, fast response, and high information storage capacity. However, current high-precision micro/nano processing methods, such as FIB (Focused Ion Beam) and semiconductor lithography, are far too complex and in lack of flexibility. Femtosecond laser, as a non-contact, high-precision, high-intensity tool for "cold" processing, is particularly favored in micro/nano processing. This review first elucidated the background and mechanism of femtosecond laser micromachining used in optical device. After that, we discussed a number of methods employed to improve the resolution of femtosecond micromachining. Then we listed various advanced processing means based on femtosecond laser and systematically summarized recent representative research developments of femtosecond laser micromachining used in microlens, gratings, optical waveguides, and photonic crystals. Finally, we concluded the challenges and the directions for further development of femtosecond laser machining in the field of micro-optical devices.

Femtosecond laser two-photon polymerization (TPP) micro-nano fabrication technology, as an important method for the preparation of three-dimensional (3D) micro-nanostructures, has become a hot spot of international frontier research. Using the two-photon absorption effect and the threshold effect of the interaction between laser and matter, this technology can break through the diffraction limit of classical optical theory and achieve nanoscale laser fabrication resolution. It is expected to play an important role in the field of 3D functional micro-nano device fabrication. In this paper, the basic principles of photophysical and photochemical processes in femtosecond pulsed laser TPP fabrication technology will be described, and the research progress and development of this technology in improving line width and fabrication resolution, and improving fabrication efficiency will be reviewed. Then, using the high spatial resolution and true 3D fabrication characteristics of femtosecond laser TPP micro-nano fabrication technology, the researchers prepared various micro-optical devices, integrated optical devices, micro-electromechanical systems, and biomedical devices, fully demonstrating the application prospect of this technology. Finally, how to achieve high-precision, high-efficiency, low-cost, large-area, multi-functional materials and microstructure fabrication, as well as existing challenges and future development directions are discussed and prospected.

Microstructure sensor is a sensor device with the micro-scale structure as a sensitive unit and can convert external physical, chemical, and biological signals into electrical signals. It has been widely used in intelligent robots, health monitoring, virtual electronics, and other fields. At present, the manufacturing methods of microstructure sensors mainly include laser manufacturing technology, MEMS technology, and 3D printing technology. Laser manufacturing technology is a green processing method that focuses the high-energy laser beam on the object to be processed and makes the laser interact with the material, mainly including laser ablation, laser direct writing, laser induction, and laser-template composite processing. It has the advantages of non-contact processing, no mask, and customizable manufacturing. By optimizing the parameters of the laser manufacturing process, it can realize the efficient and low-cost manufacturing of microstructures with different sizes and shapes. In this paper, the types, function, and manufacturing technology of the microstructures are summarized. At the same time, the microstructure sensors fabricated by laser manufacturing technology are summarized and classified, and the manufacturing technology and application of bioelectric sensors, temperature sensors and pressure sensors are analyzed in detail. Finally, the development trends of the laser manufacturing technology for microstructure sensors are summarized and prospected.

Overview: The mid-infrared (MIR) wavelength coincides with various molecular resonances and spectroscopy, which is universally used to identify chemical and biological substances. In particular, the 13 μm~20 μm wavelength window has fingerprints of unique material groups such as organometallic, halogenated, and aromatic bonds. Thus, the MIR supercontinuum generation (SCG) is widely used in the fields of biomedicine, spectroscopy, and environmental science. Thanks to the mature semiconductor growth technology combined with the advanced CMOS integration technology, SCG in on-chip devices has been studied in recent years. Cadmium telluride (CdTe) has an ultra-broad transparent spectral range, from 0.86 μm to 25 μm, and one of the largest third-order nonlinear coefficients (n2~5×10?17 m2/W at 1.55 μm, 2×10?17 m2/W at 5.5 μm, which are several times larger than that of silicon) among the MIR materials, which makes CdTe become an excellent candidate for long-wavelength MIR on-chip SCG. As an important material of solar cells, there is a mature film growth and etching technology for CdTe. In this work, we designed a large-core CdTe integrated waveguide on a low-refractive-index cadmium sulfide (CdS) film with a silicon substrate. The waveguide structure is designed with CdS as the intermediate cladding layer to achieve a low waveguide loss and high mode confinement. A large-core CdTe waveguide is tailored to generate a low and flat dispersion in the MIR spectral range, while balancing the large effective nonlinearity and the convenience of coupling. The effective refractive index in the CdTe waveguide is obtained by finite element method. Then, the simulated results solved by the nonlinear Schr?dinger equation manifest that a CdTe waveguide with a propagation distance as short as 1 cm can broaden the MIR spectrum covering 4.1 μm to 9.7 μm pumped by a 5.5 μm femtosecond laser. Experimentally, polycrystalline CdS and CdTe films were deposited by magnetron sputtering, and the CdTe waveguides were fabricated by photolithography followed by wet etching. In particular, the sidewall of the waveguide is almost perpendicular to the substrate due to the large difference in the longitudinal and transverse etching rates caused by the unique grain arrangement of the film. A near-infrared femtosecond laser centered at 1030 nm with a pulse width of 250 fs at a 500 kHz repetition rate is employed as the pump source, and an apparent spectral broadening based on self-phase modulation was observed. The numerical simulations match well with the experimental results. These results pave the way for long-wavelength mid-infrared light sources and provide abundant new opportunities for MIR microphotonics.The mid-infrared (MIR) wavelength coincides with various molecular resonances and spectroscopy. It is a universal way to identify chemical and biological substances. Thus, the MIR supercontinuum generation (SCG) is widely used in biomedicine, spectroscopy, and environmental science. Cadmium telluride (CdTe) has an ultra-broad transparent spectral range, from 0.86 μm to 25 μm, and one of the largest third-order nonlinear coefficients. It makes CdTe become an excellent candidate for long-wavelength MIR on-chip SCG. As an important material of solar cells, there is a well-established thin film growth technology for CdTe. We designed a CdTe integrated waveguide on a low-refractive-index CdS film with a silicon substrate. The simulation results solved by the nonlinear Schr?dinger equation manifest that the MIR SCG covering 4.1 μm to 9.7 μm can be generated from a 1 cm CdTe waveguide pumped by a 5.5 μm femtosecond laser. We experimentally fabricated the waveguide via the lithography and wet-etching techniques. The spectral broadening based on self-phase modulation from the large-core CdTe integrated waveguide is demonstrated by a femtosecond laser at the central wavelength of 1030 nm with a pulse width of 250 fs. The numerical simulations match well with the experimental results. These results pave the way for long-wavelength mid-infrared light sources and provide abundant new opportunities for MIR micro photonics.

Overview: The environment perception ability of the rover is the basis of its intelligent movement and detection, and obstacle detection and recognition is an important aspect of the environment perception, and the recognition effect directly determines the work ability and safety of the rover. At present, the obstacle recognition of Mars exploration vehicles mainly relies on binocular cameras. This passive measurement method based on vision is easy to fail in 3D reconstruction in weak texture and low brightness areas. As a direct measurement method, lidar has better performance in the face of the above disadvantage scenarios, so it has attracted more attention in the current hot field of automatic driving. This paper proposes an automatic obstacle recognition method for the Mars surface based on lidar data. Firstly, based on the analysis of the laser reflection intensity theory, the point cloud intensity was corrected according to the distance and angle factors through the intensity compensation theory, so as to eliminate the intensity difference of homogeneous ground objects caused by the difference in distances and angles, and then the reflection relationship between the laser radar intensity value and the target feature was accurately constructed. The global threshold was automatically obtained by the Otsu method, and the point cloud on the Mars surface was adaptively classified into an obstacle point cloud and a non-obstacle point cloud. Then, the curvature threshold is set, the unqualified obstacle point cloud is eliminated by curvature constraint, and the obtained point cloud belongs to the obstacle. Finally, the connectivity clustering based on octree leaf nodes is used to segment the obstacle point cloud into independent individuals. On this basis, the typical obstacles larger than a specific size are separated from the obstacle point cloud by setting the obstacle diameter size threshold, so as to realize the automatic recognition of the Martian surface obstacle point cloud. The size of the simulated Martian surface field tested in this paper is 22 m×16 m, and the main obstacles in the scene are rocks and other vehicle detectors. The experimental data collection and processing of the simulated field show that the proposed method can effectively extract the Martian surface obstacles in the lidar point cloud, and the recognition accuracy of typical obstacles is close to 90%, which can provide a reference for the related research based on the Martian rover obstacle detection and environmental perception. Of course, the current popular deep learning method is also a highly intelligent recognition method, so the obstacle point cloud recognition based on deep learning is also a kind of idea worthy of subsequent discussion and experiment.The environment perception ability of the Mars rover is the basis of its intelligent movement and detection. Obstacle detection is an important aspect of environment perception, which directly determines the working ability and safety of the Mars rover. In this paper, a method of identifying obstacles on the surface of Mars based on LiDAR data is proposed. Based on the obtained LiDAR point cloud data, the intensity of the point cloud is modified according to the distance and angle factors through the intensity compensation theory based on the analysis of the laser reflection intensity theory, and then the reflection relationship between the lidar intensity value and the target characteristics is constructed. The global threshold is automatically obtained through the Otsu method, and the Mars surface point cloud is adaptively classified into an obstacle point cloud and a non-obstacle point cloud. Then, the obstacle point cloud which does not meet the conditions is removed by curvature constraint. Finally, using the connectivity clustering based on Octree-based leaf nodes, the recognition of the obstacle point cloud on the surface of Mars is realized. Through the simulation experiment, the results show that this method can effectively extract the obstacles on the surface of Mars from the LiDAR point cloud, and provide a reference for the related research based on the obstacle monitoring of the Mars rover and environmental perception.

Overview: Lithography accuracy is one of the key parameters to measure the lithography process. Lithography projection objective is the core component of a lithography machine. The distortion of the lithography projection objective is one of the most important factors that affect the overlay accuracy. It causes the position offset exposed on the silicon wafer from the ideal position to the actual position. At present, the detection techniques for the distortion of the projection objective are exposure measurement, aerial image measurement, and wavefront measurement. The exposure measurement that is suitable for lithography manufacturers depends on the exposure process and equipment, thus its detection process is complex. The repeatability of exposure measurement is better than 1 nm. Both the aerial image measurement and the wavefront measurement without exposure are based on the photoelectric sensors. Its detection speed is fast, and the measurement accuracy can be better than 0.4 nm. The image quality detection platform of the projection objective is a device for detecting the optical performance of the projection objective. The wavefront sensor is a device for measuring wavefront aberrations and can also be used for distortion measurement. According to the principle of distortion measurement, when the Hartmann sensor is used to detect the distortion of the projection objective, the measurement accuracy of the mask's actual imaging position through the objective directly affects the accuracy of the distortion detection. So this paper focuses on the analysis of the displacement measurement errors of the image quality detection platform when using the Shack-Hartmann sensor to detect the distortion of the projection objective. The factors that affect the displacement measurement accuracy of the image quality detection stage mainly include the following aspects: the measurement error of the dual-frequency laser interferometer, the horizontal Abbe error of the distortion measurement, and the reproducibility error of the Hartmann position measurement. The measurement errors of the dual-frequency laser interferometer include the instrument error, the geometric error, and the environmental error. These errors are related to the performance of the interferometer, the layout of the interferometer, the performance of the image quality detection stage, and the environmental conditions. The horizontal Abbe error in distortion measurement is due to the fact that the dual-frequency laser interferometer cannot accurately measure the position coordinates of other points outside the central field in distortion detection. The reproducibility error of the Hartmann position measurement means the reproducibility of the Hartman sensor position information measured aiming at a same objective in a period of time. The factors that affect the reproducibility error of the Hartmann position measurement include the fluctuation of the interferometer value, the position control accuracy of the motion stage, the long-term drift of the interferometer, the long-term stability of the Hartmann sensor, the longterm stability of the test light source, the optical properties stability of the objective, etc. In this paper, a set of projection objective image quality detection platform is taken as an example to analyze the errors of its displacement measurement. Its displacement measurement error is about 22 nm within the measurement range of 40 mm × 40 mm. At the same time, the distortion of a projection objective is detected by this image quality detection platform, and the measurement result is about 80 nm. The displacement measurement error in the distortion detection of the lithographic projection objective is one of the important error sources of distortion detection. Depth analysis of the error sources and reduction of the error terms can improve the distortion detection accuracy.In the distortion detection of the lithography projection objective, the displacement measurement error is one of the important error sources. Depth analysis of the error sources and reduction of the error terms can improve the distortion detection accuracy. Combining the positioning and measurement technology of the moving stage, this paper analyzes the displacement measurement error of the image quality detection stage when the Shack-Hartmann wavefront sensor is used to detect the distortion of the projection objective. In this paper, a set of image quality detection platform is taken as an example to analyze the displacement measurement error in the distortion detection of the projection objective, and the image quality detection platform is used to measure the distortion of a projection objective. The distortion detection result is about 80 nm, in which the displacement measurement error of the image quality detection platform will bring about an uncertainty of about 22 nm to the distortion detection result.

Total generalized variation is effective and widely used in natural image denoising and deblurring due to its ability to suppress the staircase effect while preserving image edges and details. In order to improve the reconstruction performance of blind deconvolution on solar images, total generalized variation and PSF regularization are introduced into the reconstruction of solar images. A space-invariant multi-frame blind deconvolution model via second-order total generalized variation is proposed in this paper to improve the robustness of noise and recover more texture details. The model is solved by alternating minimization of the image sub-model and the PSF sub-model, where the image submodel can be solved by the half-quadratic splitting method. Combined with the non-blind deconvolution based on hyper-Laplacian prior, a space-invariant multi-frame blind deconvolution algorithm can be established under the multiscale framework. Then, by overlapping image segmentation and weighted stitching, the space-invariant blind deconvolution algorithm is extended to a reconstruction algorithm suitable for wide field-of-view solar images, which can reduce reconstruction errors caused by anisoplanatism. Finally, the reconstruction experiment and analysis are carried out on the real solar images observed by the one-meter New Vacuum Solar Telescope (NVST) in southwest China. The results show that the algorithm has good image reconstruction performance in both subjective visual effects and objective indexes. Second-order total generalized variation regularization and multi-frame can improve the stability and reliability of solar image reconstruction.Blind deconvolution is one of the commonly used post-reconstruction methods for adaptive optics images. In order to improve the reconstruction performance of blind deconvolution on solar (adaptive optics) images, a space-variant multi-frame blind deconvolution model based on second-order total generalized variation is proposed. It first solves the proposed space-invariant blind deconvolution model via second-order total generalized variation by the alternating minimization and half-quadratic splitting method. Then, according to the characteristics of wide field-of-view solar images which are anisoplanatic, the space-variant in the proposed algorithm is implemented by overlapping image segmentation and weighted stitching. Finally, the reconstruction experiment and analysis are carried out on the real solar images observed by the one-meter New Vacuum Solar Telescope (NVST). The results show that the proposed algorithm has good image reconstruction performance in both subjective visual effects and objective indexes.

Therefore, aiming at the problem of long-distance complex noise interference and signal attenuation, combined with the characteristics that variance analysis can detect the vibration quickly and effectively in one-dimensional signal, this paper proposes a dynamic variance threshold algorithm, and uses the parallel programming technology to improve the response speed of the system. The signal preprocessed by the band-pass filter is processed by variance processing, Gaussian blur, threshold peak seeking, and accurate center of gravity. The problem of long response time caused by the attenuation of Rayleigh scattering signal and the large amount of computation in the long-distance DVS detection is solved. The parallel programming technology is used to improve the operation speed by 184%, so as to quickly and accurately determine the location of the disturbance.The difference between the man-made disturbance and the noise on a 39 km long optical fiber is experimentally studied, and the influence of the noise is eliminated by the dynamic variance algorithm.The response time of the system is 1 second, the spatial resolution is 20 meters, and the positioning error is less than 0.1%.In order to solve the problems of weak positioning accuracy, low sensitivity, and slow response speed of the distributed fiber vibration sensor system, a dynamic variance threshold algorithm based on the phase-sensitive photosensitive time domain reflection is proposed. The signal preprocessed by the band-pass filter is processed by variance processing, Gaussian blur, threshold peak seeking, and accurate center of gravity. The problem of long response time caused by the attenuation of Rayleigh scattering signal and the large amount of computation in the long-distance DVS detection is solved. The parallel programming technology is used to improve the operation speed by 184%, so as to quickly and accurately determine the location of the disturbance. The difference between the man-made disturbance and the noise on a 39 km long optical fiber is experimentally studied, and the influence of the noise is eliminated by the dynamic variance algorithm. The response time of the system is 1 second, the spatial resolution is 20 meters, and the positioning error is less than 0.1%.

The experiments are carried out from the aspects of validity verification of each network module, comparison of subjective visual effects, comparison of quantitative evaluation results, and algorithm complexity. The performance of the proposed network is verified on five public light field data sets. The proposed algorithm obtains high-resolution light field sub-aperture images with higher PSNR and SSIM.Based on the advanced imaging technology, light field camera can obtain the spatial information and the angular information of the scene synchronously. It achieves higher dimensional scene representation by sacrificing the spatial resolution. In order to improve the spatial resolution of the light field image, a light field super-resolution reconstruction network based on angle difference enhancement is built in this paper. In the proposed network, eight multi-branch residual blocks are used to extract shallow features. Then, four enhanced angular deformable alignment modules are used to extract deep features. Finally six simplified residual feature distillation modules and pixel shuffle modules are used to complete data reconstruction. The proposed network takes advantage of the angle difference of the light field to complete the spatial information super-resolution. In order to obtain more features difference between different views, the own feature of the single view is emphasized during the feature extraction. The performance of the proposed network is verified on five public light field data sets. The proposed algorithm obtains high-resolution light field sub-aperture images with higher PSNR and SSIM.

Overview: Abundant resources in the ocean are not explored yet. The meaning of ocean exploration is far-reaching for the development of human society. In the future, the construction of underwater observation networks with outstanding performance is the precondition of a variety of scientific experiments, and time-frequency networks like GPS are helpful for the collaborative working of underwater platforms. The success of underwater wireless optical communication expands the dissemination of time-frequency signals over free-space and fiber links to underwater links. Compared with the conventional underwater frequency transfer methods, i.e. sonar, fiber links, and microwave method, the laser-based underwater frequency transfer owns strong competitiveness with the features of high flexibility, high bandwidth, and low latency. Such advantages make the laser-based underwater frequency transfer a promising approach for the construction of future underwater time-frequency networks. The paper introduces the progress on laser-based underwater frequency transfer at the University of Electronic Science and Technology of China, including the property analysis of underwater links, three specific techniques, and the future works that would be conducted. The property analysis of underwater links was conducted from the timing fluctuation attributed to refractive-index perturbation and underwater turbulence introduced power spectral density (PSD) that derived from the Kolmogorov model. The simulation results of Kolmogorov PSD and its modified PSD (von Kármán model) are given and analyzed. Based on the analysis, the experimental demonstrations of frequency transfer over 3 m, 6 m, and 9 m underwater links were conducted. The experimental results show that timing fluctuation is most partly attributed to underwater turbulence. It is necessary for highly stable frequency transfer to suppress timing fluctuation. An electronic phase compensation technique was employed for timing fluctuation suppression. A 100 MHz radio-frequency (RF) signal has been transferred over a 5 m underwater link using this technique. With the help of this technique, the root-mean-square (RMS) timing fluctuation was successfully suppressed from 9.6 ps to 2.1 ps within 5000 s. The noise, limited compensation bandwidth, and residual timing fluctuations of the electronic phase shifters block further improvement of system performance. Consequently, the optical delay line-based optical phase compensation technique was proposed and experimentally demonstrated. A 500 MHz RF signal was transferred over a 5 m underwater link for 5000 s, the RMS timing fluctuation was successfully suppressed from 7.3 ps to 0.162 ps. The experimental results show that the proposed technique could effectively suppress timing fluctuation and lower the system noise floor. Both phase compensation techniques are hard to support multiple-access frequency transfer because compensation configuration is included in the transmitter. A novel scheme for multiple-access frequency transfer was proposed and experimentally demonstrated. A 100 MHz RF signal was transferred over a 3 m underwater link for 5000 s, the RMS timing fluctuation was successfully suppressed from 73.4 ps to 3 ps. With this scheme, the phase of the frequency signal at multiple receivers could be simultaneously locked to the phase of the reference signal at the transmitter. In the future, the optical link of the multipleaccess frequency transfer experimental setup would be optimized for further performance improvement. Laser-based underwater frequency transfer experiments with picosecond-level timing fluctuations over hundred meters links would be demonstrated with the help of the single-photon detection technique and a picosecond mode-locked laser at 1064 nm wavelength.Inspired by underwater wireless optical communication, laser-based underwater frequency transfer technology extends frequency transfer and dissemination from fiber links and free-space links to underwater links and shows greater potential for applications. Compared with traditional underwater frequency transfer technologies (sonar, fiber links), laser-based underwater frequency transfer technology is more flexible and avoids the multipath effect and high latency. In the future, this technology is expected to contribute to the applications of underwater navigation and sensing, distributed observation networks, tracking and positioning systems, etc. This paper first introduces the background and significance of the underwater laser-based frequency transfer technique, and briefly shows the achievements of domestic and foreign scientific research institutions in underwater laser-based frequency transfer. Next, the paper presents the time domain and frequency domain descriptions of underwater link properties, in which the former is based on the refractive index perturbation of the water column and the latter is based on the Kolmogorov atmospheric turbulence model. Then, the research results of the University of Electronic Science and Technology in laser-based underwater frequency transfer are reported, including the electrical phase compensation technique, the optical phase compensation technique, and the multiple-access frequency dissemination technique. Finally, the three laser-based underwater frequency transfer experiments are summarized, and the future works of our group in laser-based underwater frequency transfer have been prospected. As a promising underwater frequency transfer technology, laser-based underwater frequency transfer technology will play a crucial role in relevant applications in the future.

After testing with 50 frames of random data, the experimental results show that the clustering accuracy and clustering time of this method are 86.96% and 23 ms, respectively, which are better than other algorithms, and can be used in mobile robot navigation and obstacle avoidance, autonomous handling, and other fields.Aiming at the requirement of fast clustering and segmentation of 3D point clouds for mobile robots in the process of perception of unknown environments, a 3D laser point cloud clustering method based on image information constraints is proposed. Firstly, the effective 3D environment information is obtained through point cloud preprocessing, and the RANSAC method is used to segment and eliminate the ground point cloud. Secondly, the sensor data is introduced into the YOLOv5 target detection algorithm after completing the spatiotemporal registration, and the K-means clustering algorithm of the 3D point cloud is improved. The detection frame range of the 2D image target is used to constrain the 3D point cloud and reduce the interference of non-target objects. The parameter initialization of the point cloud clustering algorithm is realized based on the image detection information. The clustering results are optimized by the intra-class outlier elimination method. Finally, the mobile robot hardware platform is built, and the box is tested. The experimental results show that the clustering accuracy and clustering time of the method in this paper are 86.96% and 23 ms, respectively, which can be used in mobile robot navigation and obstacle avoidance, autonomous handling, and other fields.

Overview: The phase-sensitive optical time-domain reflectometry (Φ-OTDR) sensing system has the characteristics of high dynamic response and high sensitivity, and has great application potential in the field of large-scale engineering structural health monitoring. The instrumentation level and engineering application of Φ-OTDR systems depend to a large extent on digital signal processing (DSP) technology. For the Φ-OTDR system, the tasks of digital signal processing mainly include three aspects. First, the demodulation of Rayleigh's backscattered light phase information should be completed accurately and efficiently. It is necessary to understand the relationship between the phase difference and the sound field signal. Then, it is necessary to reasonably set the core parameters of the Φ-OTDR system in the digital-to-analog conversion to obtain the RBS signal quickly and accurately. After that, it is necessary to select an appropriate demodulation method for demodulation. Second, all kinds of noise floor of the sensing system itself should be analyzed and suppressed. Since the noise floor of the sensing system itself is inevitable, analyzing and suppressing it is the key to improve the signal-to-noise ratio of the system. The drift of the laser center frequency, the local birefringence change of the fiber, and the nonlinear correspondence between the fiber strain and the interference intensity will all introduce corresponding noise to the system. Among the many types of noise, the coherent fading brought by the system will cause the system SNR to continue to deteriorate and randomly form detection blind spots; the polarization-related noise caused by the external environment will affect the Φ-OTDR system's ability to perceive multiple disturbance events. Third, reliable feature extraction and pattern recognition strategies should be quickly selected to improve the accuracy and intelligence of system reconstruction disturbance events. In engineering applications, various monitoring objects and time-varying background noise make it difficult to describe vibration events by accurate mathematical models. In particular, when Φ-OTDR is used in new scenarios, it needs to be able to quickly establish a corresponding analysis model based on industry knowledge, and minimize the degree of manual participation in it. Therefore, efficient and reliable object feature extraction methods, pattern recognition algorithms, and machine learning strategies are urgently needed. In view of the above problems, this paper summarizes the main digital signal processing methods and technologies of the Φ-OTDR system in recent years in the digitization of optoelectronic signals, the demodulation of phase information, the suppression of system noise, and the pattern recognition of detected objects. Two application cases of transmission line condition monitoring and buried cable breakage early warning illustrate the digital signal processing skills in the design of engineering application schemes.The phase-sensitive optical time-domain reflectometry (Φ-OTDR) sensing system has the characteristics of high dynamic response and high sensitivity, and has great potential in the field of large-scale engineering structural health monitoring. The instrumentation level and engineering application of Φ-OTDR systems depend to a large extent on digital signal processing (DSP) technology. This paper compares and analyzes the main digital signal processing methods and technologies of Φ-OTDR systems in signal quantization, demodulation, noise suppression, and pattern recognition in recent years. The importance and method of combining digital signal processing with industry background knowledge in engineering applications are expounded, and the development status and trend of the digital signal processing methods in Φ-OTDR systems are summarized and prospected.

A method for preparing microlens arrays based on projection lithography was proposed, and microlens arrays of various calibers and different surface roughness were successfully prepared by the method. The method employs a 0.2× projection objective lens to reduce the manufacturing cost of masks and realize the preparation of microlens arrays with different calibers. We achieve superior surface figure accuracy while reducing the complexity of mask preparation by employing a projection-based mask-shift filtering technique. Four kinds of microlens arrays with different calibers, 50 μm, 100 μm, 300 μm and 500 μm, were prepared. The machining accuracy of the surface morphology reaches the sub-micron level and the surface roughness reaches the nanometer level. The experimental results show that this method has great potential in the fabrication of microlens arrays, and can achieve lower line width and higher surface profile accuracy than traditional methods.

There are several major challenges in the detection and identification of contraband in millimetre-wave synthetic aperture radar (SAR) security imaging: the complexities of small target sizes, partially occluded targets and overlap between multiple targets, which are not conducive to the accurate identification of contraband. To address these problems, a contraband detection method based on dual branch multiscale fusion network (DBMFnet) is proposed. The overall architecture of the DBMFnet follows the encoder-decoder framework. In the encoder stage, a dual-branch parallel feature extraction network (DBPFEN) is proposed to enhance the feature extraction. In the decoder stage, a multi-scale fusion module (MSFM) is proposed to enhance the detection ability of the targets. The experimental results show that the proposed method outperforms the existing semantic segmentation methods in the mean intersection over union (mIoU) and reduces the incidence of missed and error detection of targets.

Unsupervised person re-identification has attracted more and more attention due to its extensive practical application prospects. Most clustering-based contrastive learning methods treat each cluster as a pseudo-identity class, overlooking intra-class variances caused by differences in camera styles. While some methods have introduced camera-aware contrastive learning by partitioning a single cluster into multiple sub-clusters based on camera views, they are susceptible to misguidance from noisy pseudo-labels. To address this issue, we first refine pseudo-labels by leveraging the similarity between instances in the feature space, using a weighted combination of the nearest neighboring predicted labels and the original clustering results. Subsequently, it dynamically associates instances with possible category centers based on refined pseudo-labels while eliminating potential false negative samples. This method enhances the selection mechanism for positive and negative samples in camera-aware contrastive learning, effectively mitigating the influence of noisy pseudo-labels on the contrastive learning task. On Market-1501, MSMT17 and Personx datasets, mAP/Rank-1 reached 85.2%/94.4%, 44.3%/74.1% and 88.7%/95.9%.

In order to meet the requirements of miniaturization, compact structure and high resolution of the space optical systems in the fields of Earth remote sensing observation and spaceborne Lidar detection, this paper designs a compact off-axis triple inverse system based on Zernike free-form surface, which simultaneously meets the requirements of long focal length, small distortion and wide working band. The system adopts an off-axis triple inverse optical system with the asymmetric and nearly circular layout, and the third mirror of the system adopts the free-form surface design. By setting appropriate optimization objectives and methods in Zemax software, the design of the optical system is optimized. Finally, the effective focal length of the system is 800 mm, the F-number is 4, the field of view is 12°×6°, and the distortion is less than 1%. The working band covers the visible and near/middle infrared bands, and the ground element resolutions of 1.5 m (visible light) and 2.5 m (near infrared) can be achieved at the orbit height of 400 km, and the ground width is 80 km×40 km. The analysis and verification of system aberration, dot plot, MTF and other performance indexes are carried out. The results show that the design scheme brings the high resolution and improves the information acquisition ability.

Aiming at the problems of poor low-light image quality, noise, and blurred texture, a low-light enhancement network (DF-DFANet) based on dual-frequency domain feature aggregation is proposed. Firstly, a spectral illumination estimation module (FDIEM) is constructed to realize cross-domain feature extraction, which can adjust the frequency domain feature map to suppress noise signals through conjugate symmetric constraints and improve the multi-scale fusion efficiency by layer-by-layer fusion to expand the range of the feature map. Secondly, the multispectral dual attention module (MSAM) is designed to focus on the local frequency characteristics of the image, and pay attention to the detailed information of the image through the wavelet domain space and channel attention mechanism. Finally, the dual-domain feature aggregation module (DDFAM) is proposed to fuse the feature information of the Fourier domain and the wavelet domain, and use the activation function to calculate the adaptive adjustment weight to achieve pixel-level image enhancement and combine the Fourier domain global information to improve the fusion effect. The experimental results show that the PSNR of the proposed network on the LOL dataset reaches 24.3714 and the SSIM reaches 0.8937. Compared with the comparison network, the proposed network enhancement effect is more natural.

Due to the difficulties of complex backgrounds and large-scale differences between objects during the process of ship multi-object tracking in sea-surface scenarios, an improved CSTrack algorithm for ship multi-object tracking is proposed in this paper. Firstly, as violent decoupling is used in the CSTrack algorithm to decompose neck features and cause object feature loss, an improved cross-correlation decoupling network that combines the Res2net module (RES_CCN) is proposed, and thus more fine-grained features can be obtained. Secondly, to improve the tracking performance of multi-class ships, the decoupled design of the detection head network is used to predict the class, confidence, and position of objects, respectively. Finally, the MOT2016 dataset is used for the ablation experiment to verify the effectiveness of the proposed module. When tested on the Singapore maritime dataset, the multiple object tracking accuracy of the proposed algorithm is improved by 8.4% and the identification F1 score is increased by 3.1%, which are better than those of the ByteTrack and other algorithms. The proposed algorithm has the advantages of high tracking accuracy and low error detection rate and is suitable for ship multi-object tracking in sea-surface scenarios.

Far-field super-resolution microscopic imaging technology based on fluorescent labels opened a gate to the microscopic world, which has become an important tool in the research of modern medicine and life science. However, the development of far-field unlabeled super-resolution microscopy is relatively slow. Here, an integrated differential microscopic imaging method using optical fiber devices is proposed in this article. The generation of hollow spots in the differential imaging system is realized by a special fiber mode selection coupler (MSC). The problem of strict alignment between hollow and solid spots is naturally solved in this method. A highly integrated label-free microscopic imaging system was established. In experiments, gold particles with a diameter of 150 nm and unlabeled polymer lines with a minimum spacing of about 50 nm were imaged to test the imaging system. The resolution of the imaging system shows great improvement compared to conventional scanning confocal microscopy.

The point ahead angle mechanism (PAAM) is a key component of the space gravitational wave detection telescope. It can control the displacement of the telescope precisely by inputting voltage or charge to the piezoelectric actuator. Therefore, the displacement response of the piezoelectric ceramic actuator directly affects the pointing control performance of the PAAM. In this paper, the equivalent capacitance calculation method is proposed to quantitatively analyze the displacement response characteristics of piezoelectric actuators driven by charge, and the accuracy and feasibility of the calculation method are verified by numerical simulation and experimental verification. The results show that when a charge amplifier controlled by 5 V, 0.05 Hz~5 Hz sine wave signal is used to drive a certain type of piezoelectric actuator, the maximum deviation of displacement response between the analysis results and the experimental results is within 1.35%, which provides a possible analysis method and realization way for the high-precision pointing control of the PAAM of the space gravitational wave detection telescope.

In gravitational-wave detection systems, the surface scattering properties of ultra-smooth optical components play a crucial role in achieving high-precision gravitational-wave measurements. To analyze and predict the surface scattering properties of ultra-smooth optical components accurately and rapidly, a non-paraxial scalar scattering model, the Generalized Beckmann-Kirchhoff (GBK) model, was built up. On this basis, the influences of both the incident angle and the scattering azimuth angle on the angular resolved scattering distributions of both P-polarized and S-polarized incident light were investigated. Under different statistical distribution characteristics of optical surfaces, the effects of incident angle, azimuth angle, autocorrelation length, slope, cut-off frequency, and surface roughness on the scattering angle resolution distribution were analyzed. The research results can provide useful references for the manufacturing of ultra-smooth optical components and the generation and mitigation of stray light in gravitational-wave detection systems.

The optical telescopes for space-based gravitational wave missions play an important role in the measurement, which both expand the beam going to the far spacecraft and efficiently collect the beam sent from the far spacecraft. The telescope, as part of the interferometric path, directly affects the measurement noise. Compared with the imaging system, the telescope for space gravitational wave observatory not only has high requirements on wavefront quality, but also has extremely high requirements on stray light performance and optical path stability, and the latter two are more challenging. The research progress of the telescope's optical system, optical-mechanical structure, space environment and thermal design, stray light simulation and suppression, and stability measurement is reviewed, which can provide a reference for the development of space gravitational telescope in our country.

Accurate measurement and control of wavefront aberrations in space-based telescopes are key to achieving efficient space gravitational wave detection. This paper presents a method for measuring wavefront aberrations of space-based telescopes based on the Shack-Hartmann wavefront sensor. This method employs a cross-correlation algorithm in the frequency domain after frequency domain threshold denoising. The measurement accuracy of the algorithm is verified using a Shack-Hartmann wavefront sensor with 20×16 sub-apertures, microlens dimensions of 0.279 mm×0.279 mm, and a focal length of 34 mm. Point source images with known defocus RMS values (0, 0.22, 0.44, and 0.66 nm) are generated, producing point source images with displacements. After wavefront reconstruction using the modal method, the RMS values of the reconstructed and residual wavefronts are calculated, comparing the measurement accuracy of the cross-correlation algorithm in the frequency domain with the traditional centroid algorithm. The results show that as the actual defocus value increases, the measurement error of the centroid algorithm presents an upward trend, respectively at 0.0966 nm, 0.1378 nm, 0.1284 nm, and 0.1463 nm. The cross-correlation algorithm in the frequency domain can increase the measurement accuracy by 13%, 7%, 18%, and 14% respectively, providing an important reference for the high-precision testing of wavefront aberrations of space gravitational wave space-based telescopes on the ground.

The telescopes for space-based gravitational wave detection are used to transmit the laser beam between spacecraft to support the precise interference measurement system. Therefore, the optical path stability of the telescope has become a crucial technical parameter. In this system, pupil aberrations provide deeper insights compared to traditional image plane aberrations in understanding optical path stability requirements, evaluating telescope imaging quality, and suppressing tilt-to-length coupling noise. Based on the theory of traditional imaging aberration and pupil aberration theory, the initial structure of the telescope is established, and the automatic correction of pupil aberration and image plane aberration is achieved through macro programming in the commercial optical software Zemax, thus achieving the design of a high-performance spaceborne telescope. Simulation results demonstrate that the design can meets the requirements of the TianQin mission.

In the space gravitational wave detection system, the accuracy of the complex amplitude field distribution simulation at the telescope exit pupil closely affects the accuracy of the interferometric measurement and impacts the effectiveness of TTL noise control analysis. Therefore, it is necessary to carry out the high-precision diffraction calculation for the field propagation. This paper demonstrates the necessity of the vectorial ray-based diffraction integral algorithm for simulation and illustrates the algorithm flow combining the telescope model. A computational model was established based on the algorithm. The telescope system parameters were substituted to verify the wavefront calculation accuracy, and the vectorial field simulation results were presented. Based on the system model, the effects of the input Gaussian field parameters and the complex refractive index on the output vectorial optical field characteristics are simulated.

Gravitational wave detection imposes high stability requirements on telescopes in space. To achieve independent measurement and calibration of the optical path stability accuracy of the telescope, the research was conducted on corresponding measurement methods. Based on the principle of heterodyne interferometric measurement, a high common-mode suppression interferometric measurement scheme was designed, and an optical path noise theoretical model was established. According to the requirement of 1 pm/Hz1/2@1 mHz for optical path stability indicators, the optical path noise level of the measurement system components was allocated. To verify the feasibility of the scheme and the accuracy of the noise theoretical model, an interferometric measurement system was constructed at the front end of the telescope. According to the relevant parameters of the experimental instruments and optical components, the theoretical evaluation of the system's optical path noise level was 7.319 nm/Hz1/2@10 mHz. The experimental measurement result of 3 nm/Hz1/2@10 mHz was consistent with the theoretical evaluation, indicating that the interferometric path has good noise common-mode suppression characteristics, and verifying the accuracy of the noise theoretical model. When the testing environment and instrument accuracy meet the requirements for optical path noise allocation, this measurement scheme is expected to achieve the measurement of the optical path stability of gravitational wave telescope.

This paper focuses on the ultra-low thermal deformation requirements of the main support structure of the gravitational wave detection telescope. It proposes a method to reduce the thermal deformation of the truss support structure by designing CFRP (carbon fiber reinforced polymer) layers to modify the material's thermal expansion coefficient. Additionally, to meet the alignment performance requirements of the telescope, a segmented design scheme for the structure is presented. The paper begins by analyzing the advantages of CFRP, existing methods of thermal dissipation, and the research progress both domestically and internationally. It determines the three-segment telescope design using CFRP as the support material and establishes design criteria. Next, mathematical models for "material-thermal deformation" and "truss structure-thermal deformation" are developed. Optimization is conducted for material layering and structural design, resulting in an optimized solution. Furthermore, CFRP materials are applied to the support structure, and a segmented main support structure design scheme is proposed to reduce the difficulty of structural processing and alignment. The overall structure is analyzed. The analysis results demonstrate that, in terms of mechanical performance, the overall structure's natural frequency and maximum gravity-unloading deformation meet the requirements of the main support structure. In terms of thermal deformation, the optimized design based on CFRP layering exhibits a thermal deformation that is 27.15% of the conventional layering scheme, 6.42% of the Invar material support rod scheme, 11.50% of the SiC support rod scheme, and 3.21% of the titanium alloy support rod scheme. This indicates that the optimized design can significantly reduce the structural thermal deformation.

To address the challenges of complex models, slow detection speed, and high false detection rate during fire detection, a lightweight fire detection algorithm is proposed based on cascading sparse query mechanism, called LFNet. In the study, firstly, a lightweight feature extraction module ECDNet is established to extract more fine-grained features in different levels of feature layers by embedding the lightweight attention module ECA (efficient channel attention) in YOLOv5s backbone network, which is used to solve the multi-scale of flame and smoke in fire detection. Secondly, deep feature extraction module FPN+PAN is adopted to improve multi-scale fusion of feature maps at different levels. Finally, the Cascade Sparse Query embedded lightweight cascade sparse query module is applied to improve the detection accuracy of small flames and thin smoke in early fires. Experimental results show that the comprehensive performance of the proposed method in objective indicators such as mAP and Precision is the best on SF-dataset, D-fire and FIRESENSE. Furthermore, the proposed model achieves lower parameters and higher detection accuracy, which can meet the fire detection requirements of challenge scenes.

Single-image rain removal is a crucial task in computer vision, aiming to eliminate rain streaks from rainy images and generate high-quality rain-free images. Current deep learning-based multi-scale rain removal algorithms face challenges in capturing details at different scales and neglecting information complementarity among scales, which can lead to image distortion and incomplete rain streak removal. To address these issues, this paper proposes an image rain removal network based on cross-scale attention fusion, aiming to remove dense rain streaks while preserving original image details to improve the visual quality of the rain removal image. The rain removal network consists of three sub-networks, each dedicated to obtaining rain pattern information at different scales. Each sub-network is composed of densely connected cross-scale feature fusion modules. The designed module takes the cross-scale attention fusion as the core, which establishes inter-scale relationships to achieve information complementarity and enables the network to consider both details and global information. Experimental results demonstrate the effectiveness of the proposed model on synthetic datasets Rain200H and Rain200L. The peak signal-to-noise ratio (PSNR) of the derained images reaches 29.91/39.23 dB, and the structural similarity index (SSIM) is 0.92/0.99, outperforming general mainstream methods and achieving favorable visual effects while preserving image details and ensuring natural rain removal.

In order to suppress the time-varying disturbance in the tip-tilt correction system, an adaptive disturbance suppression method based on characteristic disturbance frequency identification is proposed in this paper. The least mean square error criterion is used to identify the characteristic disturbance frequency of the closed-loop system error, and the identified filtering parameters and controller adjustment are designed in parallel. At the same time, a method based on frequency splitting is proposed: combining low-frequency disturbance with high-frequency disturbance suppression, which further improves the speed of characteristic frequency identification and simplifies the design process, and realizes adaptive disturbance suppression in the closed-loop bandwidth. The closed-loop verification of the proposed method is carried out in the tip-tilt correction device. The experimental results show that the method can quickly identify the characteristic disturbance and adaptively adjust the controller, which can improve the closed-loop performance of the system under single-frequency and multi-frequency time-varying disturbance.

For the sparse angle projection data, the problem of artifact and noise is easy to appear in the image reconstruction of computed tomography, which is difficult to meet the requirements of industrial and medical diagnosis. In this paper, a sparse angle CT iterative reconstruction algorithm based on overlapping group sparsity and hyper-Laplacian prior is proposed. The overlapping group sparsity reflects the sparsity of image gradient, and the overlapping cross relation between the adjacent elements is considered from the perspective of the image gradient. The hyper-Laplacian prior can accurately approximate the heavy-tailed distribution of the image gradient and improve the overall quality of the reconstructed image. The algorithm model proposed in this paper uses alternating direction multiplier method, principal component minimization method and gradient descent method to solve the objective function. The experimental results show that under the condition of the sparse angle CT reconstruction, the proposed algorithm has certain improvement in preserving structural details and suppressing noise and staircase artifacts generated in the process of image reconstruction.

To address the problem that existing networks find it difficult to learn local geometric information of point cloud effectively, a graph convolutional network that fuses multi-resolution features of point cloud is proposed. First, the local graph structure of the point cloud is constructed by the k-nearest neighbor algorithm to better represent the local geometric structure of the point cloud. Second, a parallel channel branch is proposed based on the farthest point sampling algorithm, which obtains point clouds with different resolutions by downsampling them and then groups them. To overcome the sparse characteristics of the point cloud, a geometric mapping module is proposed to perform normalization operations on the grouped point cloud. Finally, a feature fusion module is proposed to aggregate graph features and multi-resolution features to obtain global features more effectively. Experiments are evaluated using ModelNet40, ScanObjectNN, and ShapeNet Part datasets. The experimental results show that the proposed network has state-of-the-art classification and segmentation performance.

An adaptive feature fusion cascaded Transformer retinal vessel segmentation algorithm is proposed in this paper to address issues such as pathological artifacts interference, incomplete segmentation of small vessels, and low contrast between vascular foreground and non-vascular background. Firstly, image preprocessing is performed through contrast-limited histogram equalization and Gamma correction to enhance vascular texture features. Secondly, an adaptive enhancing attention module is designed in the encoding part to reduce computational redundancy while eliminating noise in retinal background images. Furthermore, a cascaded ensemble Transformer module is introduced at the bottom of the encoding-decoding structure to establish dependencies between long and short-distance vascular features. Lastly, a gate-controlled feature fusion module is introduced in the decoding part to achieve semantic fusion between encoding and decoding, enhancing the smoothness of retinal vessel segmentation. Validation on public datasets DRIVE, CHASE_DB1, and STARE yielded accuracy rates of 97.09%, 97.60%, and 97.57%, sensitivity rates of 80.38%, 81.05%, and 80.32%, and specificity rates of 98.69%, 98.71%, and 98.99%, respectively. Experimental results indicate that the overall performance of this algorithm surpasses that of most existing state-of-the-art methods and holds potential value in the diagnosis of clinical ophthalmic diseases.

There are noises and speckles in OCT retinal images, and a single extraction of spatial features is often easy to miss some important information. Therefore, the target region cannot be accurately segmented. OCT images themselves have spectral frequency domain characteristics. Aiming at the frequency domain characteristics of OCT images, this paper proposes a new dual encoder model based on U-Net and fast Fourier convolution to improve the segmentation performance of the retinal layer and liquid in OCT images. The proposed frequency encoder can extract image frequency domain information and convert it into spatial information through fast Fourier convolution. The lack of feature information that can be omitted by a single space encoder will be well-complemented. After comparison with other classical models and ablation experiments, the results show that with the addition of a frequency domain encoder, the model can effectively improve the segmentation performance of the retinal layer and liquid. Both average Dice coefficient and mIoU are increased by 2% compared with U-Net. They are increased by 8% and 4% compared with ReLayNet, respectively. Among them, the improvement of liquid segszmentation is particularly obvious, and the Dice coefficient is increased by 10% compared with the U-Net model.

Conventional optical imaging is essentially a process of recording and reproducing the intensity signal of a scene in the spatial dimension with direct uniform sampling. Therefore, the resolution and information content of imaging are inevitably constrained by several physical limitations, such as optical diffraction limit and spatial bandwidth product of the imaging system. How to break these physical limitations and obtain higher resolution and broader image field of view has been an eternal topic in this field. Computational optical imaging, by combining front-end optical modulation with back-end signal processing, offers a new approach to surpassing the diffraction limit of imaging systems and realizing super-resolution imaging. In this paper, we introduce the relevant research efforts on improving imaging resolution and expanding the spatial bandwidth product through computational optical synthetic aperture imaging, including the basic theory and technologies based on coherent active synthetic aperture imaging and incoherent passive synthetic aperture imaging. Furthermore, this paper reveals the pressing demand for "incoherent, passive, and beyond-diffraction-limit" imaging, identifies the bottlenecks, and provides an outlook on future research directions and potential technical approaches to address these challenges.

In IDRID dataset, the sensitivity and specificity were 95.65% and 91.17%, respectively, and the quadratic weighted agreement test coefficient was 90.38%. In the Kaggle competition dataset, the accuracy rate is 84.41%, and the area under the receiver operating characteristic curve was 90.36%. The experimental results show that the algorithm in this paper has certain application value in the field of DR. In view of the shortcomings of the above model, the next key task is to streamline the network model and further improve the model performance as much as possible.Diabetic Retinopathy (DR) is a prevalent acute stage of diabetes mellitus that causes vision-effecting abnormalities on the retina. In view of the difficulty in identifying the lesion area in retinal fundus images and the low grading efficiency, this paper proposes an algorithm based on multi-feature fusion of attention mechanism to diagnose and grade DR. Firstly, morphological preprocessing such as Gaussian filtering is applied to the input image to improve the feature contrast of the fundus image. Secondly, the ResNeSt50 residual network is used as the backbone of the model, and a multi-scale feature enhancement module is introduced to enhance the feature of the lesion area of ??the retinopathy image to improve the classification accuracy. Then, the graphic feature fusion module is used to fuse the enhanced local features of the main output. Finally, a weighted loss function combining center loss and focal loss is used to further improve the classification effect. In the Indian Diabetic Retinopathy (IDRID) dataset, the sensitivity and specificity were 95.65% and 91.17%, respectively, and the quadratic weighted agreement test coefficient was 90.38%. In the Kaggle competition dataset, the accuracy rate is 84.41%, and the area under the receiver operating characteristic curve was 90.36%. Simulation experiments show that the proposed algorithm has certain application value in the grading of diabetic retinopathy.

Overview: The success of the deep detection model largely requires a large amount of data for training. Under the condition of fewer training samples, the model is easy to overfit and the detection effect is unsatisfactory. In view of the model that is easy to overfit and cause the target misdetection and missed detection in the absence of training samples, we present the Few-Shot Object Detection via the Online Inferential Calibration (FSOIC) framework by using the Faster R-CNN as detector. Through its excellent detection performance and powerful ability to distinguish the foreground and background, it effectively solves the problem that the single-stage detector cannot locate the target when the training samples are scarce. The bottom-layer features have a larger size and stronger location information, but the lack of global vision leads to weak semantic information, while the top-layer features are the opposite. To make full use of the sample information, the framework is designed to possess a new Attention-FPN network, which selectively the fuses features through modeling the dependencies between the feature channels, and directs the RPN module to extract the correct novel classes of the foreground objects by combined with the hierarchical freezing learning mechanism. The channel attention mechanism compresses the feature map and spreads it into a one-dimensional vector for sigmoid through two fully connected layers. The weight is generated for each feature channel, and the correlation between each channel is established. The weight of the input features is allocated according to the category, and the dependence relationship between each channel is modeled. Due to the closed nature of the neural network, simple feature fusion is uncertain, and it is difficult to fuse the feature map in a satisfactory direction. To the imbalanced sample features, the candidate targets of the new class are scored too low and filtered in the selection of the prediction box, resulting in false detection and missed detection of the detector. We designed the online calibration module that segmentes and encodes the samples, scored the re-weighted the multiple candidate objects, and corrected the misdetected and missed predicted objects. The performance of our detection algorithm performs better than most comparisons. The experimental results in the VOC Novel Set 1 show that the proposed method improves the average nAP50 of the five tasks by 10.16% and performs better than most comparisons.Considering that the model is easy to overfit and cause the target misdetection and missed detection under the condition of few samples, this paper propose the few-shot object detection via the online inferential calibration (FSOIC) based on the two-stage fine-tuning approach (TFA). In this framework, a novel Attention-FPN network is designed to selectively fuse the features by modeling the dependencies between the feature channels, and direct the RPN module to extract the correct novel classes of the foreground objects in combination with the hierarchical freezing learning mechanism. At the same time, the online calibration module is constructed to encode and segment the samples, reweight the scores of multiple candidate objects, and correct misclassifying and missing objects. The experimental results in the VOC Novel Set 1 show that the proposed method improves the average nAP50 of the five tasks by 10.16% and performs better than most comparisons.

Overview: Breast cancer is the most common cancer among women. Tumor grading based on microscopic imaging is important for the diagnosis and prognosis of breast cancer, and the results need to be highly accurate and interpretable. Breast cancer tumor grading relies on pathologists to assess the morphological status of tissues and cells in the microscopic images of tissue sections, such as tissue differentiation, nuclear isotypes, and mitotic counts. A strong statistical correlation between the hematoxylin-Eosin (HE) stained microscopic imaging samples and the progesterone Receptor (ER) immunohistochemically (IHC) stained microscopic imaging samples has been documented, i.e., the ER status is strongly correlated with the tumor tissue grading. Therefore, it is a meaningful task to use deep learning models to research the breast tumor grading in ER IHC pathology images exploratively. At present, the CNN module integrating attention has a strong ability of induction bias but poor interpretability, while the Vision Transformer (ViT) block-based deep network has better interpretability but weaker ability of induction bias. In this paper, we propose an end-to-end deep network with adaptive model fusion by fusing ViT blocks and CNN blocks with integrated attention. Due to the negative fusion phenomenon of the existing model fusion methods, while it is impossible to guarantee that ViT blocks and CNN blocks with integrated attention have good feature representation capability at the same time; in addition, the high similarity and redundant information between the two feature representations lead to a poor model fusion capability. To this end, this paper proposes an adaptive model fusion method that includes multi-objective optimization, adaptive feature representation metric, and adaptive feature fusion, which effectively improves the fusion ability of the model. The accuracy of the model is 95.14%, which is 9.73% better than that of ViT-B/16 and 7.6% better than that of FABNet. The visualization of the model focuses more on the regions of nuclear heterogeneity (e.g., giant nuclei, polymorphic nuclei, multinuclei and dark nuclei), which is more consistent with the regions of interest to pathologists. Overall, the proposed model outperforms the current state-of-the-art model in terms of accuracy and interpretability.Tumor grading based on microscopic imaging is critical for the diagnosis and prognosis of breast cancer, which demands excellent accuracy and interpretability. Deep networks with CNN blocks combined with attention currently offer better induction bias capabilities but low interpretability. In comparison, the deep network based on ViT blocks has stronger interpretability but less induction bias capabilities. To that end, we present an end-to-end adaptive model fusion for deep networks that combine ViT and CNN blocks with integrated attention. However, the existing model fusion methods suffer from negative fusion. Because there is no guarantee that both the ViT blocks and the CNN blocks with integrated attention have acceptable feature representation capabilities, and secondly, the great similarity between the two feature representations results in a lot of redundant information, resulting in a poor model fusion capability. For that purpose, the adaptive model fusion approach suggested in this study consists of multi-objective optimization, an adaptive feature representation metric, and adaptive feature fusion, thereby significantly boosting the model's fusion capabilities. The accuracy of this model is 95.14%, which is 9.73% better than that of ViT-B/16, and 7.6% better than that of FABNet; secondly, the visualization map of our model is more focused on the regions of nuclear heterogeneity (e.g., mega nuclei, polymorphic nuclei, multi-nuclei, and dark nuclei), which is more consistent with the regions of interest to pathologists. Overall, the proposed model outperforms other state-of-the-art models in terms of accuracy and interpretability.

Overview: Organic photodetectors (OPDs) have advantages such as wide material sources, tunable spectrum, solution processing, and low manufacturing cost. They have numerous potential applications in domains of aviation, military, business, medicine, et al. Traditional organic photodetectors owe a planar structure and mostly use indium tin oxide (ITO), silver, aluminum, and other materials as electrodes. The organic photosensitive layer is sandwiched between two asymmetric electrodes to form a "sandwich" structure. However, the rigidity and brittleness of planar substrates constrain their application in flexible and wearable devices. As an inherently flexible and simple-to-weave material, fibers have been widely used in electronic textiles and wearable devices in recent years. For example, there has been substantial research on fibrous solar cells, supercapacitors, light-emitting devices, and physical/chemical sensors. In terms of fiber-based photodetectors. In 2014, Jixun Chen et al used a Ni wire as the core to prepare a NiO-ZnO heterojunction, and a twine Pt wire as the outer electrode to realize ultraviolet detection. In 2018, Xiaojie Xu et al. modified CuZnS: TiO2 array on the Ti wire surface, and wrapped a carbon nanotube fiber (CNT) on the outer layer to realize the collection and transmission of photogenerated carriers. However, most fiber-based photodetectors are based on inorganic photosensitive materials, showing the disadvantages of complicated manufacturing procedures, poor flexibility, and high cost. In this study, a fiber-based organic photodetector (FOPD) was prepared using organic photosensitive material. The electron transport layer (ZnO), organic heterojunction photosensitive layer (PBDB-T:ITICTh), and hole transport layer (PEDOT:PSS) were prepared on the surface of zinc wire by dip-coating method. Silver wire or carbon nanotube fiber (CNT) was wrapped as the external electrode, and two kinds of flexible FOPDs were obtained. Both the two devices showed a typical response in the visible band with remarkable rectification characteristics, and exhibited a specific detection rate of 1011 Jones (300 nm~760 nm) at ?0.5 V bias. Due to the better interface contact between CNT external electrode and photosensitive layer, the CNT-FOPD showed a lower dark current density (9.5×10?8 A cm?2, ?0.5 V) and faster response speed (rise and fall time: 0.88 ms and 6.00 ms). This work is expected to provide new ideas for the development of flexible fibrous devices and wearable electronics.Fiber-based photodetectors are expected to be widely used in the field of wearable electronics due to their properties of flexibility, easy-to-weave, and omnidirectional light detection. Currently reported fiber-based photodetectors mostly use inorganic photosensitive materials, which have drawbacks such as limited mechanical flexibility and complex preparation processes. In this paper, we proposed the fiber-based organic photodetector (FOPD). The electron transport layer (ZnO), organic heterojunction photosensitive layer (PBDB-T:ITIC-Th), and hole transport layer (PEDOT: PSS) were prepared on zinc wire by a solution dip-coating method layer by layer. Finally, silver wire or carbon nanotube fiber (CNT) was wrapped as the external electrode, and two kinds of flexible FOPDs were obtained and showed typical rectification characteristics. They showed a specific detection rate of 1011 Jones (300 nm~760 nm) at ?0.5 V bias. Due to the better interface contact between the CNT external electrode and photosensitive layer, the CNT-based device exhibited lower dark current density (9.5×10?8 A cm?2, ?0.5 V) and faster response speed (rise time of 0.88 ms and fall time of 6.00 ms). The work is expected to provide new ideas for the development of flexible fiber-based devices and wearable electronics.

Overview: Intravascular ultrasound (IVUS) is an important imaging modality for diagnosing cardiovascular diseases. The annotation of major anatomical structures of blood vessels in IVUS images can provide necessary clinical parameters for lesion severity assessment, which is a necessary step for physicians' diagnosis. However, manual annotation is laborious and inefficient. With the development of deep learning, convolutional neural networks perform well in this task and are able to achieve automatic and accurate segmentation and recognition of the main anatomical structures of blood vessels. Existing IVUS image segmentation networks are mostly based on pixel-by-pixel prediction, which lacks overall constraints on the main structures of blood vessels and cannot guarantee that the topological relationships between main vessel structures conform to medical prior knowledge, which has a negative impact on the calculation of clinical parameters. To solve this problem, this paper proposes an IVUS image segmentation method based on polar coordinate modeling and a dense-distance regression network. First, a prior knowledge-based polar coordinate modeling is designed for encoding the two-dimensional mask of the main structure of blood vessels containing prior knowledge into a one-dimensional distance vector to avoid the topological relationship of the blood vessel structure from generating random changes in the network prediction. A dense-distance regression network consisting of a residual network and a semantic embedding branching module is then constructed for learning the mapping relationship between IVUS images and 1D distance vectors. To effectively constrain the learning direction of the network, a joint loss function is proposed. This loss function takes into account the actual spatial relationship between one-dimensional distance vectors and has a stronger supervisory capability. The network prediction results are finally reconstructed as a two-dimensional mask by spline curve fitting. The proposed method is validated on a 20 MHz IVUS image dataset. The experimental results show that the proposed method achieves 100% topology preservation in the media, lumen, and plaque regions and achieves the Jaccard measure (JM) of 0.89, 0.87, and 0.74, respectively. The advantage of the algorithm in this paper is that it can provide a high accuracy and topologically correct segmentation results of the vessel structures, which is suitable for general IVUS image segmentation. The clinical parameters provided are reliable and can be used as an important reference basis for physicians' diagnosis, reducing physicians' workload and improving diagnostic efficiency, which has a promising future in clinical applications.Aiming at the problem that existing intravascular ultrasound (IVUS) image segmentation networks cannot guarantee that the topological relationships between segmentation results conform to medical prior knowledge, which has a negative impact on clinical parameter calculation, an IVUS image segmentation method based on polar coordinate modeling and dense-distance regression network is proposed. This method converts two-dimensional (2D) masks to one-dimensional (1D) distance vectors to preserve the topology of the vessel structures through polar coordinate modeling with prior knowledge. Then a dense-distance regression network consisting of a residual network and semantic embedding branch is constructed for learning the mapping relationships between IVUS images and 1D distance vectors. A joint loss function is proposed to constrain the network learning direction. The prediction results are finally reconstructed as 2D masks by spline curve fitting. The experimental results show that the proposed method achieves 100% topology preservation in the media, lumen, and plaque regions, and achieves Jaccard measure (JM) of 0.89, 0.87, and 0.74, respectively. The algorithm is suitable for general IVUS image segmentation, with high accuracy, and can provide reliable clinical parameters.

Overview: The state of retinal blood vessels is an important indicator for clinicians in the auxiliary diagnosis of eye diseases and systemic diseases. In particular, the degree of atrophy and pathological conditions of retinal blood vessels are the key indicators for judging the severity of the diseases. Automatic segmentation of retinal blood vessels is an indispensable step to obtain the key information. Good segmentation results are conducive to accurate diagnosis of the eye diseases. Due to the good characteristic of U-Net that can use skip connection to connect multi-scale feature maps, it performs well in segmentation tasks with small data volume, therefore, it could be applied to retinal vascular segmentation. However, U-Net ignores the features of retinal blood vessels in the training process, resulting in the inability to fully extract the feature information of blood vessels, while its segmentation results show that the vessel pixels are missing or the background noise is incorrectly segmented into blood vessels. Researchers have made various improvements on U-Net for the retinal vessel segmentation task, but the methods still ignore the global structure information and boundary information of retinal vessels. To solve the above problems, a boundary attention assisted dynamic graph convolution retinal vessel segmentation model based on U-Net is proposed in this paper, which supplements the model with more sufficient global structure information and blood vessel boundary information, and extracts more blood vessel feature information as much as possible. First, RGB image graying, contrast-limited adaptive histogram equalization, and gamma correction were used to preprocess the retinal images, which can improve the contrast between the vascular pixels and background, and even improve the brightness of some vascular areas. Then, rotation and slice were adopted to enhance the data. The processed images were input into the model to obtain the segmentation result. In the model, dynamic graph convolution was embedded into the decoder of U-Net to form multiscale structures to fuse the structural information of feature maps with different scales. The method not only can enhance the ability of dynamic graph convolution to obtain global structural information but also can reduce the interference degree of the noise and the segmenting incorrectly background on the vascular pixels. At the same time, in order to strengthen the diluted vascular boundary information in the process of up-down sampling, the boundary attention network was utilized to enhance the model’s attention to the boundary information for the sake of improving the segmentation performance. The presented model was tested on the retinal image datasets, DRIVE, CHASEDB1, and STARE. The experimental results show that the AUC of the algorithm on DRIVE, CHASEDB1 and STARE are 0.9851, 0.9856 and 0.9834, respectively. It is proved that the model is effective.Aiming at the problem of missing and disconnected capillary segmentation in the retinal vascular segmentation task, from the perspective of maximizing the use of retinal vascular feature information, by adding the global structure information and retinal blood vessels boundary information, based on the U-shaped network, a dynamic graph convolution for retinal vascular segmentation model assisted by boundary attention is proposed. The dynamic graph convolution is first embedded into the U-shaped network to form a multi-scale structure, which improves the ability of the model to obtain the global structural information, and thus improving the segmentation quality. Then, the boundary attention network is utilized to assist the model to increase the attention to the boundary information, and further improve the segmentation performance. The proposed algorithm is tested on three retinal image datasets, DRIVE, CHASEDB1, and STARE, and good segmentation results are obtained. The experimental results show that the model can better distinguish the noise and capillary, and segment retinal blood vessels with more complete structure, which has generalization and robustness.

The significance of OCT imaging in ophthalmic surgery has been proved in experiments on animal eyes, human eye models, and clinical cases. In recent years, commercial OCT surgical navigation equipment has already been widely used in ophthalmic clinical surgery. With the progress of image processing technology, image and ophthalmology, OCT surgical navigation equipment will further promote the innovation of ophthalmic surgery, and thus promote the development of the ophthalmology field.During ophthalmic microsurgery, the visualization of internal structures is limited by traditional intraoperative imaging methods due to their lack of depth information. Optical coherence tomography (OCT) is a non-contact tomographic imaging technique that is widely used for intraoperative navigation in ophthalmic surgery because of its ability to provide depth information, non-invasiveness, fast imaging, and high resolution. Typical OCT devices can be divided into handheld OCT and microscope-integrated OCT. This article briefly introduces the mechanism and development of time domain OCT and fourier domain OCT, reviews the development of OCT ophthalmic surgical navigation devices, introduces representative OCT systems in each category, describes and compares their imaging principles, performance, advantages, and disadvantages, and finally concludes with a summary and outlook on the applications of this technology in ophthalmic surgery.

In recent years, slightly off-axis digital holography which combines the advantages of off-axis and on-axis has been vigorously developed. In order to further improve its real-time performance, a synchronous slightly off-axis system based on the field of view (FOV) multiplexing technique has been applied. However, the spatial position of the holograms collected by this technology is unknown, which causes a spatial mismatch problem. In order to ensure the accuracy of the subsequent holographic reconstruction, it is necessary to perform a spatial mismatch calibration. The existing calibration methods can be roughly divided into: intensity-based calibration methods and phase-based calibration methods. Intensity-based calibration methods are susceptible to environmental noise, and phase-based calibration methods only have pixel-level accuracy. At the same time, none of the existing methods take into account the longitudinal position error caused by the sensor tilt.The existing spatial mismatch calibration methods can only achieve pixel-level calibration accuracy due to the limitation of principle, and are easily disturbed by the environmental noise. In this paper, a spatial mismatch calibration method for a sub-pixel-level simultaneous slightly off-axis digital holographic microscope system is proposed. In the modeling, the method not only analyzes the lateral position error caused by the image segmentation, but also further considers the longitudinal position error caused by the sensor tilt, and summarizes the calibration process as a nonlinear multi-variable optimization problem. In this paper, the particle swarm optimization algorithm is used to solve this optimization problem because of its simple structure, high convergence efficiency, and strong global search ability. In the calibration process, a phase-only wavefront based on the phase aberration is established, and the root mean square error of the phase-only wavefront is used as the target function to remove the influence of noise on the calibration accuracy. Simulation results show that the proposed method has sub-pixel accuracy, and experiment demonstrates the effectiveness of the method in the practical systems.

Overview: With the advent of the 5G communication era, much attention has been paid to free manipulate electromagnetic waves at a subwavelength scale. Meta-surface with subwavelength structural dimensions have shown broad prospects in the field of microelectronic components due to their powerful electromagnetic control capabilities. In this paper, a subwavelength comb-shaped space-filling meta-surface is designed by using metal curves according to the resonator principle. A series of studies on spoof localized surface plasmon resonance characteristics are carried out on this basis. Theoretical analysis and calculation are carried out according to the structural characteristics. Compared with the traditional meta-surface supporting spoof localized surface plasmons, this curved arrangement of continuous metals will form an air waveguide similar to a resonant cavity, allowing for larger waveguide lengths at smaller dimensions, resulting in greatly reduced working frequency band. Under the excitation of the incident electromagnetic wave, spoof localized surface plasmon like Fabry-Perot resonance will be generated. The resonance frequency of the meta-surface can be calculated from the resonance conditions. Using the finite element method to simulate the 2D comb structure with different periods, it is found that the Q-factor of 1.7×105 can be obtained when the structure compression ratio (λ/L) is 444 by adjusting the structure period. In the study of the higher-order eigenmodes of the comb-shaped space-filled meta-structure, it was found that the spoof localized surface plasmons excited by space-filling structures are alternately supported by magnetic and electric multipoles modes, and the scattering cross-section of the eigenmodes of each order are presented at equally spaced frequencies. By changing the distribution type of the space-filling structure, the supported surface plasmon resonance properties are not affected by the arbitrary bending of the structure, and the magnetic field intensity distribution of the eigenmodes only changes with the direction of the air waveguide. Finally, the 3D simulation of the comb-shaped space-filling structure is carried out, from the X-Z section electric field diagram, it can be observed that the spoof localized surface plasmons generated by the structure can bind the energy on the surface of the structure and generate localized field enhancement. The space-filling design in this paper makes full use of the structure space. This highly localized structure can generate a higher Q-factor under the deep subwavelength structure, and the electromagnetic properties are not affected by the arbitrary bending of the metal structure, and have better stability. It provides a new idea for the preparation of nanometer-sized high-efficiency electromagnetic resonators. With the advent of the 5G communication era, much attention has been paid to manipulate electromagnetic waves at subwavelength scale. In this paper, we propose comb-shaped meta-structures based on space-filling curves, and use theoretical analysis and numerical simulation method to study the near-field electromagnetic properties of these meta-surfaces. Finally, the effective excitation of all order eigenmodes of spoof localized surface plasmon resonance can be realized in these meta-structures. Through adjusting the structure period to change the effective length of the air waveguide, high compression ratio between resonant wavelength and device size, and high Q-factor can be simultaneously achieved. Moreover, spoof localized surface plasmons excited by space-filling structures are alternately supported by magnetic dipole and electric dipole modes. As a consequence, changing the distribution form of the space-filling curve with the remaining parameters unchanged, the resonant characteristics of the surface plasmon supported by the structure are not affected by the shape tortuosity, but are only related to the total length of the equivalent waveguide. Thus, the space-filling curvilinear structure can be freely designed. We believe that, our results have great potential in designing the high Q-factor miniaturized electromagnetic resonator devices based on spacing-filling curvilinear meta-structure.

This paper is verified on the Apollo data set. The continuous road live screenshots are extracted from the data set to obtain the required set of time-series pictures, and the targets in the images are detected and tracked. Finally, the experiments show that the algorithm in this paper has an obvious effect on solving the occlusion problem. This paper has also been tested on the actual road, and the effect of medium and long-distance vehicle detection is good. The experiment shows that the algorithm can meet the real-time detection requirements under the actual road conditions.In the field of automatic driving target tracking, there is a problem that the target occlusion will cause the loss of feature points, resulting in the loss of tracking targets. In this paper, a multi-target tracking algorithm combining spatial mask prediction and point cloud projection is proposed to reduce the adverse effects of the occlusion. Firstly, the temporal image data is processed by an example segmentation mask extraction model, and the basic mask data is obtained. Secondly, the obtained mask data is input into the tracker, the mask output of subsequent sequence images is obtained through the prediction model, and the verifier is used for a comparative analysis to obtain an accurate target tracking output. Finally, the obtained 2D target tracking data is projected into the corresponding point cloud image to obtain the final 3D target tracking point cloud image. In this paper, simulation experiments are carried out on multiple data sets. The experimental results show that the tracking effect of this algorithm is better than other similar algorithms. In addition, this paper is also tested on the actual road, and the vehicle detection accuracy reaches 81.63%. The results verify that the algorithm can also meet the real-time requirements of target tracking under the actual road conditions.

In summary, the basic principle, sensing scheme, and performance of F-SBS optical fiber sensors are introduced in this paper. With the F-SBS sensor applied in practice, increasing demand for high accuracy, and high spatial resolution emerges, which we believe will be dominant in the research of substance identification sensors in the future.Forward stimulated Brillouin scattering (F-SBS), a 3-order nonlinear effect in optical fibers, has become the hotspot in recent years, due to its great potential in substance identification, and fiber diameter measurement, etc. Through research and analysis of the progress of F-SBS, the main principle and key techniques are generalized in this paper. Distributed sensing schemes based on local light phase recovery, opto-mechanical time-domain reflectometry, and opto-mechanical time-domain analysis are emphatically introduced here. With the gradual practical application of F-SBS, the demand for distributed measurement of F-SBS with high precision and high spatial resolution becomes more and more significant, which will be the main research direction of F-SBS in optical fibers in the future.

In order to make the recognition of sea fog with high accuracy and reasonable interpretability, the cloud-aerosol LiDAR with orthogonal polarization (CALIOP), which is capable of penetrating clouds and obtaining atmospheric profiles, was first used to annotate medium and high cloud, low cloud, sea fog, and clear sky sea surface samples. Then, bright temperature features and texture features were extracted for each type of sample in combination with multi-channel data from the Himawari-8 satellite. Finally, according to the needs of sea fog monitoring, the inference decision tree for sea fog monitoring was abstracted and a deep neural decision tree model was built accordingly, which achieves high accuracy for nighttime sea fog monitoring while having strong interpretability. The continuous observation data of Himawari-8 on the night of June 5, 2020 was selected to test the sea fog. The monitoring results can clearly show the dynamic development process of the sea fog events. At the same time, the proposed sea fog monitoring method has an average probability of detection (POD) of 87.32%, an average false alarm ratio (FAR) of 13.19%, and an average critical success index (CSI) of 77.36%, which provides a new method for disaster prevention and mitigation of heavy fog at sea.Remote sensing satellites have the characteristics of wide coverage and continuous observation, and are widely used in research related to the sea fog identification. Firstly, the Cloud-Aerosol LiDAR with Orthogonal Polarization (CALIOP), which is capable of penetrating clouds and obtaining atmospheric profiles, was used to annotate medium and high cloud, low cloud, sea fog, and clear sky sea surface samples. Then, bright temperature features and texture features were extracted from each type of sample in combination with multi-channel data from the Himawari-8 satellite. Finally, according to the needs of sea fog monitoring, the inference decision tree for sea fog monitoring was abstracted and a deep neural decision tree model was built accordingly, which could achieve high accuracy for nighttime sea fog monitoring while having strong interpretability. The continuous observation data of Himawari-8 on the night of June 5, 2020 was selected to test the sea fog. The monitoring results can clearly show the dynamic development process of the sea fog events. At the same time, the sea fog monitoring method in this paper has an average probability of detection (POD) of 87.32%, an average false alarm ratio (FAR) of 13.19%, and an average critical success index (CSI) of 77.36%, which provides a new method for disaster prevention and mitigation of heavy fog at sea.

Overview: This paper is devoted to the research of synthetic aperture radar (SAR) real-time imaging processor. As the number of SAR imaging channels increases, the number of SAR imaging channels also presents new challenges. The optical processor not only has strong parallel processing ability, but also has the advantages of low power consumption, small volume, fast processing speed and programmability. Therefore, this paper designs and analyzes the SAR real-time imaging optical processor from the perspective of optical mechanical system design. Firstly, the system scheme principle of optical processor based on 4f optical structure is proposed, and the filtering algorithm is described in detail according to the principle. Secondly, according to the algorithm requirements, the relevant Fourier transform lens design is completed, and the compactness of 4f optical system is further strengthened. Then, the flexible design of the lens base is carried out, and the optimal parameter model is found by using the integrated optimization method. At the same time, it meets the modular design idea, completes the corresponding optical mechanical structure design, and obtains the optical mechanical system model of the overall scheme. The specific design results obtained based on the above research methods are as follows: in the optical design process, a Fourier transform lens with an entry pupil diameter of 21 mm, a field angle of 7°, and a focal length of 172 mm is obtained, and its MTF is better than 0.57 at 55 lp/mm. And the 4f optical system whose imaging quality tends to the diffraction limit meets the Rayleigh criterion. In the process of optical mechanical structure design, the overall size of 4f optical mechanical system is 405 mm×145 mm× 92 mm, with a mass of about 2.94 kg, and its volume and mass are only 30% and 48% of that of the inclined plane optical processor with the same SAR data processing level; At the same time, the RMS value of lens surface under normal temperature 1g gravity condition is less than λ/50(λ= 532 nm), the fundamental frequency of the overall structure is greater than 100 Hz, which can fully meet the expected design goal of the processor optical mechanical system. Finally, the simulation processing of SAR data is carried out on the optical platform. According to the simulation results, it shows that the system can be suitable for airborne or spaceborne real-time processing scenes. To sum up, the 4f optical processor designed in this paper can provide a certain reference value for improving the real-time imaging processing ability of SAR. In order to further improve the real-time imaging processing ability of synthetic aperture radar (SAR) in the face of massive echo data, the optical and mechanical system of SAR real-time imaging optical processor is designed and analyzed based on 4f optical structure. Firstly, a Fourier transform lens with an entrance pupil diameter of 21 mm, a field angle of 7°, and a focal length of 172 mm is designed for the filtering algorithm, and a compact design is adopted for the 4f optical system. Then, the flexible mirror base in 4f optical mechanical structure is optimized by using the integrated optimization method, and the overall structure is modularized designed and analyzed. The results show that the imaging quality of 4f optical system tends to the diffraction limit, and the MTF of Fourier transform lens is better than 0.57 at 55 lp/mm. The RMS value of lens surface shape of 4f optical mechanical system under normal temperature 1g gravity condition is less than λ/50. The fundamental frequency of the overall structure is greater than 100 Hz. The overall size of 4f optical processor is 405 mm×145 mm×92 mm, the mass is about 2.94 kg, and its volume and mass are only 30% and 48% of those of oblique plane optical processors with the same SAR data processing level. Through data simulation, it shows that the system design meets the needs of real-time imaging on satellite or airborne.

Overview: It is a trend of development to use the Stewart platform as a means of space. But the Stewart platform, which has a vibration ability and low system bandwidth, causing the tracking accuracy to be difficult to improve. In response to this problem, many scholars have proposed a two-stage control system, which is to design the system of a system that is far higher than the Stewart platform, which is used to curb the tracking error of the Stewart platform, thereby improving the tracking accuracy of the system. Tip-tit-mirror (TTM) bandwidth is very high, available in the intensive subsystem and the Stewart platform for two-stage control. When the system has only one CCD as a detector, it is necessary to design the decoupling link to make the system stable. But for the traditional two-order structure, it is difficult to design the decoupling element as the object of the probe. Therefore, this paper analyzes the traditional two-order structure and removes the redundant control structure. This paper also obtains the new structure based on the two-stage control of the position output and makes the decoupling link becomes easy. Through theoretical analysis, the accuracy of the system is mainly due to the ability of the precision subsystem to suppress the error of the rough system. The traditional controller of the advanced subsystem is the PI controller, which is extremely limited to the accuracy of the precision. The current research on the accuracy of the TTM tracking accuracy is mainly by introducing additional hardware devices, but this increases the uncertainty of the system stability. Therefore, the design of the PI-PI controller can effectively improve the tracking accuracy while increasing the cost of the system. After theoretical analysis and experimental verification, the PI-PI controller, compared to the PI controller, can improve the suppression ability of the error of the precision subsystem in the low frequencies. The double order structure control of the system and the Stewart platform is made up of the system, which makes the tracking precision of the system significantly improved, which is better solved the problem of the Stewart platform tracking precision that has the vibration function.The Stewart platform has six degrees of freedom motion characteristics and can be used as both vibration isolation and tracking platform. However, the vibration isolation function requires low system bandwidth, while the tracking function requires high system bandwidth, which makes it difficult to achieve high precision tracking using the Stewart platform with vibration isolation function. To solve this technical problem, a high-bandwidth tilt correction system is introduced to form a two-stage control structure, so as to improve the accuracy. The traditional two-stage control needs to design decoupling link and independent measurement sensors to achieve hierarchical control. In this paper, a control method based on a single sensor is proposed to improve the traditional dual-order structure to avoid decoupling and achieve a high-precision closed-loop for the Stewart-TTM. In order to further improve the tracking accuracy of the system at low frequencies, a PI-PI controller is designed. Theoretical analysis and experimental verification show that the Stewart dual-stage control structure based on image measurement can not only meet the requirements of vibration isolation, but also achieve high-precision tracking control. Compared with the traditional PI controller, the PI-PI control proposed in the tilt correction system can effectively improve the tracking accuracy.

Overview: Microfibers tapered from conventional optical fibers with diameters ranging from hundreds of nanometers to several micrometers possess various advantages including large evanescent field, strong light confinement, high optical nonlinearity, flexible configurability, and low-loss connection to other fiberized system, which makes it an open platform for miniaturization and integration of all-fiber devices. Nowadays microfiber can be easily obtained through mature fabrication method like flame-brushing technique. On the other hand, as a fundamental opto-electronic component, optical resonators have got comprehensively researched and widely applied in the fields of optical communication, sensing, signal processing, and quantum photonics, including whispering-gallery-mode cavities like micro-ring, micro-cylinder, micro-toroid, and micro-sphere. These traditional optical resonators are fabricated through lithography which is relatively complicated. With the maturation of microfiber fabrication methods, optical resonators based on optical microfibers have been demonstrated and developed, such as microfiber loop resonators, microfiber knot resonators, and microfiber coil resonator. As an optical coupling device based on evanescent field coupling, the microfiber resonator features in low insertion loss, high Q-factor, high finesse, excellent mechanical stability, easy fabrication process, and compatibility with fiber systems, providing a broad platform for all-fiberized miniatured devices of probing and modulation. Through further integration with exterior functional materials and microfabrication techniques, a microfiber resonator can be utilized in diverse domains of sensor, filter, modulator, and fiber laser, as well as quantum photonics and nonlinear optics, realizing the ‘lab on fiber-ring’. In the field of sensing, the microfiber resonators get exploited as the refractometric sensor, concentration and humidity sensor, temperature and current sensor, mechanical pressure sensor, microfluidic sensor, magnetic field sensor, acceleration sensor, etc., where the devices exhibit high adaptability and excellent sensitivity. As to optical signal processing, the device can be used as the single wavelength or multi-wavelength filter, code-type conversion, and optical modulation. The intensity and phase of light can be tuned to a large scale within broad wavebands, and the modulation response time is also reduced to achieve high-speed modulation. Furthermore, the microfiber resonator can be used as an optical delay line or generator of second harmonic or third harmonic. When applied into fiber laser, the microfiber resonators help build the stable light source with narrow linewidth single frequency or multiwavelength laser with high uniformity. The devices integrated with metal or 2D materials also make the laser operate under conventional soliton mode-locking or dissipative four-wave-mixing mode-locking regime and output sub-picosecond pulsation, broadening the dynamics of ultrafast optics. In this article, we summarize the recent progress in the microfiber resonators research fields, covering fundamental principles and characteristics, fabrication methods, and applications of microfiber resonators.Microfibers tapered from conventional optical fibers with diameters ranging from hundreds of nanometers to several micrometers possess various advantages including large evanescent field, strong light confinement, high optical nonlinearity, flexible configurability, and low-loss connection to other fiberized systems, which makes it an open platform for miniaturization and integration of all-fiber devices. As a fundamental opto-electronic component, optical resonators have got comprehensively researched and widely applied in the fields of optical communication, sensing, signal processing, and quantum photonics. Traditional optical resonators are fabricated through lithography which is relatively complicated. With the maturation of microfiber fabrication methods, optical resonator based on optical microfibers was demonstrated and developed. As an optical coupling device based on evanescent field coupling, the microfiber resonator features in low insertion loss, high finesse, easy fabrication, and compatibility with fiber systems. It can be utilized in domains of filter, sensor, modulator, and fiber laser. In this article, we summarize the recent progress in the microfiber resonators research fields, covering fundamental characteristics, fabrication methods, and applications of microfiber resonators.

Overview: In recent years, the number of new photoelectric measurement equipment has increased rapidly, the composition has become more and more complex, the accuracy has gradually improved, and the functions have become more comprehensive. During the normal life cycle of large-scale optoelectronic measurement equipment, engineers seek to maintain the performance of the equipment with the lowest possible cost and as few personnels as possible, so the demand for research on failure prediction and diagnosis technology is increasing. The traditional on-site manual diagnosis and maintenance method requires a lot of manpower and material resources, and it takes a long time to complete a test and diagnosis. The accuracy of the diagnosis is very dependent on the familiarity and experience of the operator. Once a fault occurs, it is difficult to quantify the time for positioning and troubleshooting, which affects the combat effectiveness of the equipment. In fact, major faults that affect the performance of equipment are generally easy to repair in the early stage, but often due to incomplete detection and diagnosis methods, they cannot be detected or cannot be detected on-site in time, resulting in major faults accumulated over time. In the fault diagnosis of photoelectric measurement system, the prediction of tracking error is particularly important. CS-BP algorithm has strong self-adaptive and self-learning ability, and can obtain more reliable results without additional human intervention, so it is often used for fault diagnosis and parameter prediction of large-scale systems. Based on the BP neural network, this article uses the cuckoo algorithm to optimize the threshold and weight, and proposes a CS-BP algorithm. This essay uses the azimuth guidance, pitch guidance, azimuth encoder, pitch encoder and time data of the photoelectric measurement system to predict the tracking error. Compared with the traditional neural network algorithm, the algorithm utilizes the cuckoo's excellent feature of finding extreme values, and solves the problem that the neural network algorithm cannot obtain the optimal solution due to improper initial threshold and weight settings. The experimental results show that compared with the traditional BP neural network and the BP neural network optimized by the genetic algorithm (GA-BP), the number of iterations of the CS-BP algorithm is 21 and 60 times less, and the average relative error of the prediction is 4.85% and 1.57% lower, respectively. Therefore, CS-BP algorithm has a faster convergence speed and higher prediction accuracy, and is suitable for application in fault diagnosis of optoelectronic measurement systems.In recent years, with the increasing number and complexity of photoelectric measurement systems, the demand for fault diagnosis is also increasing. In the fault diagnosis of the photoelectric measurement system, the prediction of its tracking error is particularly important. In this paper, we propose a BP neural network algorithm optimized by the Cuckoo algorithm (CS-BP). The tracking error can be predicted by using the azimuth guidance, pitch guidance, azimuth encoder, pitch encoder and time data of the optoelectronic measurement system. Compared with the traditional neural network algorithm, this algorithm uses the excellent characteristics of Cuckoo to find the extreme value, and solves the problem that the neural network algorithm cannot get the optimal solution due to the improper setting of the initial threshold and weight. The experimental results show that, the number of iterations with CS-BP is 21 and 60 less than the traditional BP neural network and the BP neural network optimized by the genetic algorithm (GA-BP), respectively. The relative errors are 4.85% and 1.57% lower, respectively. Therefore, the CS-BP algorithm has a faster convergence speed and higher prediction accuracy, and it is suitable for fault diagnosis of photoelectric measurement system.

Overview: In the deformation measurement of moving objects with fringe projection, such as the deformation of objects in high-speed flight and the measurement of the unconstrained facial expression changes, we pursue to project as few patterns as possible and obtain as high measurement accuracy as possible. Obtaining the unwrapped phase is one of the key steps in fringe projection 3D measurement, which generally needs the assistance of other information. The common methods are spatial phase unwrapping algorithm, temporal phase unwrapping algorithm, and multi-view geometric constraint. These methods solve the phase unwrapping problem well to some extent, but they have their limitations. The spatial phase unwrapping method is difficult to deal with spatially discontinuous or isolated regions. The temporal phase unwrapping method takes a long time and requires higher hardware at the same measurement speed. The multi-view geometric constraint method reduces the measurement area and increases the complexity and cost of the whole system. In most circumstances, the initial shape of the moving objects can be obtained. According to this fact, a two-step scheme is proposed to improve the performance of dynamic 3D shape measurement. 1) The initial 3D shape of the object and the corresponding 3D coordinates of feature points in the 2D image are obtained by measuring the static object or its CAD model. 2) Carry out the 3D measurement of object motion and change. By detecting the feature points in the dynamic image, the motion parameters of the object at different times are calculated according to the corresponding relationship between 2D and 3D coordinates. Then the approximate shape of the object is estimated from the initial shape. The approximate phase of the fringe pattern at this time is calculated. Then, combined with the approximate phase and the wrapped phase of the actual fringe, the unwrapped phase is calculated, and the 3D shape of the object at that time is obtained. Compared with the temporal phase unwrapping method, the proposed scheme improves the measurement speed under the same measurement reliability. Compared with the spatial phase unwrapping method, this scheme improves the measurement reliability at the same measurement speed and is not affected by fringe discontinuity. A static and dynamic dual-mode 3D measurement system was built by using a DLP projector and high-speed camera. The 3D shape measurement of 1280×1024 points at 70 f/s is realized. The experimental results show that the scheme is feasible and has a large tolerance for the change of the object pose at adjacent times.A two-step scheme is proposed to improve the performance of dynamic 3D shape measurement according to its characteristics. 1) The initial 3D shape of the object and the corresponding 3D coordinates of feature points in the 2D image are obtained by measuring the static object or its CAD model. 2) Carry out the 3D measurement in the process of object motion and change. By detecting the feature points in the dynamic image, the motion parameters of the object at different times are calculated according to the corresponding relationship between two-dimensional and three-dimensional coordinates. Then the approximate shape of the object is estimated from the initial shape. The approximate phase of the fringe pattern at this time is calculated. Then, combined with the approximate phase and the wrapped phase of the actual fringe, the unwrapped phase is calculated, and the 3D shape of the object at that time is obtained. Compared with the temporal phase unwrapping method, the proposed scheme improves the measurement speed under the same measurement reliability. Compared with the spatial phase unwrapping method, this scheme improves the measurement reliability at the same measurement speed and is not affected by fringe discontinuity. A static and dynamic dual-mode 3D measurement system was built by using a DLP projector and high-speed camera. The 3D shape measurement of 1280×1024 points at 70 f/s is realized. The experimental results show that the scheme can measure not only the rigid moving object but also the non-rigid moving object, as long as the fringe change caused by its deformation does not exceed half a period. The proposed method has a large tolerance for the change of object pose at adjacent times as well.

Overview: Wireless optical communication refers to the technology of transmitting information in free space using light beams as carriers, which has the advantages of high bandwidth, low cost, and high security. Due to factors such as narrow signal beam and long transmission distance, it is difficult to establish and maintain a wireless optical communication link. Therefore, an acquisition, targeting, and tracking system needs to be established to prevent the communication link from being interrupted. In the wireless optical communication system, the optical components on the two platforms carrying the transmitter and the receiver are required to be coaxial in real time, and this process is usually called automatic aiming. In order to maintain the real-time aiming of the transceiver boresight of both transceivers, it is necessary to design a fast and high-precision APT system. A typical wireless optical communication APT system is shown in Figure 1. Liu Changcheng established and analyzed the simulation model in the APT system in atmospheric laser communication, and designed an automatic beam capture system; Hu Qidi designed a beacon light spot detection scheme using CCD; Yang Peisong proposed a coaxial aiming detection method, and designed the aiming control system and tracking system according to the method, and carried out field experiments; Zhao Qi designed an initial capture system and conducted a 1.3 km field experiment; Xu Wei designed a light spot detection system and proposed a corresponding image processing algorithm; Li Shiyan proposed an optical axis aiming scheme, which can effectively improve the detection accuracy and aiming accuracy of the system; Yan Xi designed a spot tracking system and conducted a 5.2 km field tracking experiment. The experimental results show that the tracking accuracy of the system can reach 5.4 μrad; Jing Yongkang designed a light spot image detection method, and conducted a 100 km laser communication experiment on this basis; Zhang Pu embedded a high-precision actuator in the APT system to achieve high-precision aiming and tracking, designed a focusing system and conducted field experiments of 10.2 km and 100 km. Liang Hanli designed an APT system that can be mounted on UAVs and conducted an airborne laser communication experiment through a simulated airborne experimental platform, and its tracking accuracy can reach 2.42 μrad; Ke Xizheng, Yang Shangjun and others proposed a fast aiming method. The method does not need to feed back the control signal from the receiving end to the transmitting end, and can complete the establishment of the uplink and the downlink at the same time. And carried out 1.3 km and 10.3 km field experiments to verify the method. This paper systematically analyzes the development and application of the APT system in wireless optical communication and introduces the research progress and achievements of Xi'an University of Technology in this field. Including the experimental analysis and verification of the performance of the designed initial capture system, compound axis control system and beam detection system Improvements have increased the effectiveness and reliability of the APT system.Optical wireless communication refers to the technology of transmitting information in free space using light waves as a carrier, which has the advantages of high bandwidth, low cost, and high security. The acquisition, pointing, and tracking (APT) system is the premise of establishing a wireless optical communication system. A simple, reliable, and dynamic APT system can overcome the impact of mechanical platform vibration and external environment changes on the wireless optical communication system. Therefore, it is necessary to conduct in-depth theoretical and experimental research on the APT system, so as to design a capture, aiming, and tracking method suitable for wireless optical communication. This paper analyzes the domestic and foreign research achievements in capturing, aiming, and tracking, and introduces the work done by Xi'an University of Technology in the field of automatic aiming. It mainly includes the progress of initial acquisition system, non-common visual axis control system, beam detection system, etc. At the same time, the field experiments of 1.3 km, 5.2 km, 10.2 km, and 100 km distance links are introduced to verify the effectiveness of the APT system. Finally, the development of APT in wireless optical communication is prospected.

In this paper, a highly sensitive optical fiber sensor based on the Vernier effect is demonstrated for gas pressure sensing. It consists of two paralleled Fabry-Perot interferometers (FPIs), which are both produced by splicing a single-mode fiber to a short segment of capillary tube, acting as sensing cavity and reference cavity, respectively. The lateral wall of the sensing FPI is drilled with a micro-channel allowing gas to flow in. Due to the small optical path difference between the two FPI, the Vernier effect is caused in the reflected spectrum of the sensor. Thus, the gas-pressure sensitivity is significantly enhanced, achieving up to ~64 pm/kPa which is ~16 times higher than that of a single FPI. Additionally, experimental results show that the sensor is insensitive to the surrounding temperature, which reduces the influence of ambient temperature on the measurement of gas pressure. The advantages of robust structure and high sensitivity of gas pressure indicate that the demonstrated sensor has a promising potential in industrial production, gas detection, and other fields.A highly sensitive optical fiber sensor based on the Vernier effect is demonstrated for gas pressure sensing. It consists of two paralleled Fabry-Perot interferometers (FPIs), which are both produced by splicing a single-mode fiber to a short segment of capillary tube, acting as sensing cavity and reference cavity, respectively. The lateral wall of the sensing FPI is drilled with a micro-channel allowing gas to flow in. Due to the small optical path difference between the two FPI, the Vernier effect is caused in the reflected spectrum of the sensor. Thus, the gas-pressure sensitivity is significantly enhanced, achieving up to ~64 pm/kPa which is ~16 times higher than that of a single FPI. Additionally, experimental results show that the sensor is insensitive to the surrounding temperature, which reduces the influence of ambient temperature on the measurement of gas pressure. The advantages of robust structure and high sensitivity of gas pressure indicate that the demonstrated sensor has a promising potential in industrial production, gas detection, and other fields.

Overview: Terahertz detector is an important device in the field of terahertz technology, and it is important to improve its sensitivity. The sensitivity of the detector can be improved in two aspects: one is to further optimize the antenna of the detector, and the other is to optimize the size of the detector and the spot size of incident terahertz wave. Due to the long wavelength of the electromagnetic wave in the terahertz band, the spot size is much larger than the effective acceptance area of the detector, which limits the effective absorption rate of the detector to the incident terahertz wave. In order to make the focus spot to be small, the lens aperture needs to be increased. At present, the commonly used method is to integrate the hyper-hemispheric silicon lens with the terahertz detector to reduce the spot size by one order of magnitude and increase the electric field energy density. However, hyper-hemispheric silicon lens is difficult to be ultra-thin and ultra-light, and is not planar, which is not conducive to the device integration, especially for array detectors. In this paper, a series of metasurface lenses for terahertz detectors are designed using sub-wavelength silicon cylinders. By tuning the diameter of the silicon cylinders, the transmission phase of the terahertz wave can be controlled from 0 to 2π with high transmission amplitude. At 1 THz, the backside integration of the designed single-surface lens with the terahertz detector can increase the electric field energy density in the core region of the THz detector to 32 times that of the incident plane wave, and reduce the focal spot to the same order of magnitude as the wavelength. Based on the feasibility of fabrication and anti-reflection considerations, we propose a two-sided metasurface lens, which further increases the energy density of the electric field to 44 times that of the incident plane wave. Compared with the traditional hyper-hemispheric silicon lenses, the size and thickness of the metasurface lens are smaller and more convenient for integration. Metasurface lenses have a great prospect for reducing the complexity of the terahertz system and improving the responsiveness of the detector, and provide a new idea for the integration and miniaturization of the terahertz device. However, the current metasurface lenses produce many side lobes after focusing, resulting in low focusing efficiency. Further research needs to further optimize the materials and unit structures of the metasurface lenses, to improve the focusing efficiency and electric field energy density.In this paper, a focusing lens for terahertz detection is designed using a metasurface composed of sub-wavelength silicon cylinders. By tuning the diameter of the silicon cylinder, the transmission phase of the THz wave is controlled from 0 to 2π. At 1 THz, the terahertz electric field energy density focused by the single-sided metasurface lens designed can be increased to 32 times that of the incident wave. After adding the anti-reflection, a double-sided metasurface lens is proposed, which is feasible in processing, increasing the electric field energy density to 44 times that of the original. Compared with the traditional hyper-hemispheric terahertz silicon lenses, our metasurface lens has the advantages of thin thickness and small volume, which is conducive to the miniaturization of the terahertz detector component and provides the possibility to realize the integration with the terahertz detector.

Overview: As a super-resolution optical imaging technology, structured light illumination technology carries an object’s high-frequency information into the optical system in the form of moiré fringes through structured illumination, breaking the diffraction limit in traditional optical imaging and improving image resolution. An incoherent self-interference digital holography based on the Michelson interferometer can accurately record an object's phase and intensity information. It has the characteristics of fast real-time, non-contact, non-marking, three-dimensional imaging, etc., and has essential research significance in biomedical imaging and materials science. In this paper, an incoherent digital holographic imaging system based on the Michelson interferometer with structured light illumination is proposed, which uses a spatial light modulator (SLM) to realize horizontal and vertical cosine grating illumination patterns to improve the lateral resolution of the imaging system. Perform simulation and verification experiments in uniform and structured light illumination mode to explore the high-resolution imaging results of the resolution target. We obtained in simulation imagings: First, the resolved minimum element of the resolution target is Group 4 element 3 (20.16 lp/mm) in Figure 3(e) under uniform light illumination. Then, the algorithm is used to modulate the resolution target to realize the structured light illumination mode. The resolved minimum resolution element of the resolution target is Group 5 element 2 (35.92 lp/mm) in Figure 4(c). We get in the verification experiments: First, use the algorithm to generate a mask with a value of 1 on the SLM to adjust the illumination mode to the uniform light illumination mode, and the resolved minimum resolution element of the resolution target is the Group elements 4 (45.25 lp/mm) in Figure 5(e). Using another algorithm to load cosine gratings of 20 lp/mm and 40 lp/mm on the SLM to adjust the illumination mode to structured light illumination mode, the resolved minimum element of the resolution target is Group 6 element 1 (64 lp/mm) and Group 6 element 4 (90.51 lp/mm) in Figure 6(a1) and Figure 6(b1). The applicability of the super-resolution imaging method based on the structured light illumination to the incoherent light self-interference digital holographic imaging system based on the Michelson interferometer is verified from the level of simulation imaging and experiments, and the resolution of the imaging system is improved. In the future, it is necessary to comprehensively consider the system performance, optimize the system structure, study more effective numerical algorithms, and realize super-resolution imaging, dynamic imaging, color imaging, etc., to obtain more excellent development space.An incoherent digital holographic imaging system based on the Michelson interferometer with structured light illumination is proposed, which uses a spatial light modulator (SLM) to realize horizontal and vertical cosine grating illumination patterns to improve the lateral resolution of the imaging system. Using MATLAB software to carry out simulation imaging and numerical reconstruction, the high-resolution reconstructed image under the system is obtained. It theoretically proves that this method can effectively improve the resolution of the incoherent digital holography system. And build the corresponding incoherent light self-interference digital holographic imaging system. By imaging the USAF1951 resolution target, further verified the applicability of the super-resolution imaging method based on structured light illumination experimentally.

Overview: In the atmospheric environment, there are many fine particles in the air, which will lead to the absorption or refraction of light and affect the normal radiation of light. In this case, the color, contrast, saturation and detail of the image captured by the camera are often seriously affected. At present, computer vision needs to realize many high-level tasks such as pedestrian recognition, automatic driving, air navigation, remote sensing and telemetry, and these high-level tasks have a high demand for image quality. Therefore, it is of great significance to carry out single image defogging to obtain higher quality images before performing high-level tasks. In recent years, single image defogging using generative adversarial networks(GAN) has become a hot research aspect. However, the traditional GAN algorithms rely on annotated datasets, which is easy to cause over-fitting of ground truth, and usually performs not well on natural images. To solve this problem, this paper designed a GAN network incorporating dark channel prior loss to defogging single image. This prior loss can influence the model prediction results in network training and correct the sparsity and skewness of the dark channel feature map. At the same time, it can definitely improve the actual defogging effect and prevent the model from over-fitting problem. In addition, this paper introduced a new method to obtain dark channel feature map, which compresses pixel values instead of minimum filtering. This method does not need to set fixed scale to extract dark channel feature map, and has good adaptability to images with different resolutions. Moreover, the implementation function of this method is a convex function, which is conducive to embedded network training and enhances the overall robustness of the algorithm. The proposed algorithm is quantitatively analyzed in the comprehensive test set SOTS and the mixed subjective test set HSTS. The peak signal-to-noise ratio (PSNR), structural similarity SSIM and BCEA Metrics are used as the final evaluation indexes. The final result shows that our algorithm can raise PSNR up to 25.35 and raise SSIM up to 0.96 on HSTS test sets. While it comes to SOTS test sets, our method achieves the result of 24.44 PSNR and 0.89 SSIM. When we use BCEA metrics to evaluate our algorithm, we achieve the result of 0.8010 e,1.6672 r and 0.0123 p. In summary, Experimental results show that the proposed algorithm performs well on real images and synthetic test sets compared with other advanced algorithms.Single image defogging using generative adversarial networks (GAN) relies on annotated datasets, which is easy to cause over-fitting of ground truth, and usually performs not well on natural images. To solve this problem, this paper designed a GAN network incorporating dark channel prior loss to defogging single image. This prior loss can influence the model prediction results in network training and correct the sparsity and skewness of the dark channel feature map. At the same time, it can definitely improve the actual defogging effect and prevent the model from over-fitting problem. In addition, in order to solve the problem that the extraction method of traditional dark channel feature has non-convex function and is difficult to be embedded into network training, this paper introduces a new extraction strategy which compresses pixel values instead of minimum filtering. The implementation function of this strategy is a convex function, which is conducive to embedded network training and enhances the overall robustness of the algorithm. Moreover, this strategy does not need to set a fixed scale to extract the dark channel feature map, and has good adaptability to images with different resolutions. Experimental results show that the proposed algorithm performs better on real images and synthetic test-sets like SOTS when compared with other sota algorithms.

Although the FMCW LiDAR measurement technology has gradually matured, further exploration and research is still needed. At present, most of the nonlinear correction methods focus on the optical system and signal processing, but we hope to solve the problem from the design of the laser itself, and make further improvements in its mechanical structure, circuit design, and temperature control to avoid subsequent complicated work. In order to achieve true intelligence, on the one hand, we need to improve the efficiency of measurement (especially three-dimensional imaging). On the other hand, we must strive to miniaturize and integrate the FMCW LiDAR measurement system to bring more convenience and wider application scenarios.In modern measurement technology, frequency modulation continuous wave LiDAR combines the advantages of traditional radar and laser interferometry and plays an important role in the fields of the large-size space precision measurement, micro-distance measurement, and three-dimensional imaging with its characteristics such as non-contact, large measurement range, high resolution, and strong anti-jamming capability. However, in practical application, the frequency modulation of the laser light source can’t be completely linear, which greatly reduces the measurement accuracy of the frequency modulation continuous wave LiDAR technology. Therefore, how to suppress the effects of the laser frequency modulation nonlinearity has become a hot research topic in the field of frequency modulation continuous wave LiDAR measurement. This paper introduces the basic principle of the frequency modulation continuous wave LiDAR, and introduces four widely used nonlinear correction methods and some special nonlinear correction methods according to the different nonlinear correction schemes of the frequency modulation, and makes summaries and prospects.

Overview: With the development of computer vision, people increasingly need to understand images, including recognizing the scenes and the human behaviors in images. The task of HOI detection is to locate humans and objects in images and infer their relationships. This requires not only locating a single object instance, but also identifying the interaction between the objects. However, machines cannot know which object humans are interacts in. Most of the existing methods solve this problem by completely pairing the people and objects. They use off-the-shelf object detectors to detect instances, but this does not meet the requirements of the HOI task. This paper proposes an object detector suitable for HOI detection based on relational reasoning, which makes use of the interactive relationship between humans and objects in the images to recommend human-object pairs, so as to reduce the occurrence of non-interactive human-object pairs as much as possible. Our method follows the two-stage detection like most works. Firstly, the interactive instance proposal network (IIPN) is used to recommend human-object pairs. The IIPN follows the pipeline of faster RCNN, but replaces the region proposal network (RPN) with the IIPN. The IIPN selects human-object pairs based on the interaction possibility between humans and objects using the visual information in the picture. It passes the message through the iterative reasoning of the graph neural networks (GNNS), only human-object pairs that include interactive relationships are selected as the IIPN’s outputs. Secondly, we design a cross-modal information fusion module (CIFM), which calculates the fusion attention according to the influence of different features on the detection results, and performs weighted fusion. This is because the existing methods simply add or splice several features such as human visual features, object visual features, and human-object spatial features in the reasoning part. The different influence degrees of various features in different actions are ignored. For example, the verbs like ride and hold in and depend more on the spatial relationships, while eat and cut in and depend more on human's postures, that is, visual features. Meanwhile, this paper believes that semantic prior knowledge is also helpful to HOI detection. For example, if we have apples in an image, the probability of predicting the human's action as eating or holding is greater than others. Finally, complete experiments are performed on two popular large-scale HOI datasets, HICO-DET and V-COCO. The experimental results show the effectiveness of the proposed method.Human-object interaction detection is to locate and identify the interactive relationship between humans and objects in an image. The challenge is that the machine cannot know which object the person is interacting in. Most existing methods try to solve this problem by matching humans and objects exactly. Different from them, this paper proposes an interactive instance proposal network based on relational reasoning to adapt to the task. Our main idea is to recommend human-object pairs by using the potential interaction relationships in the visual relationship between humans and objects. In addition, a cross-modal information fusion module is designed to fuse different context information according to its influence on the detection result, so as to improve the detection accuracy. To evaluate the proposed method, we performed sufficient experiments on two large-scale datasets: HICO-DET and V-COCO. Results show that our method achieves 19.90% and 50.3% mAP on HICO-DET and V-COCO, which are 4.5% and 2.8% higher than our baseline, respectively.

Overview: Numerous sub-aperture fiber laser array is one of the emerging technologies to build high power, high beam quality and equivalent optical large aperture. Realizing the common phase and even the fast and flexible beam deflection of array laser beam based on the precise phase control is the key to the application of the current fiber laser phased array technology. In this paper, the optical phase-controlled steering technology is combined with the fiber laser coherent combining system, and the beam steering characteristics of the numerous sub-aperture, meter-scale fiber array laser coherent combining system are studied. Aiming at the development trend of numerous sub-aperture fiber laser phased array technology, based on the 19 aperture fiber laser phased array as the basic module, the meter-scale phased array transmitting system models with 19, 133 and 703 apertures are established. Based on the principle of optical phased array, the step phase folding model is adopted to make the piston phase distribution of the beam emitted from adjacent aperture change continuously, and to realize the high-precision continuous steering in a certain range. Meanwhile, the steering limit ranges of 19, 133 and 703 aperture fiber laser phased arrays are defined and calculated according to the distribution characteristics of the far-field steering beam pattern. Through numerical simulation analysis, the results show that when the piston phase difference of adjacent sub-apertures changes at equal intervals, the far-field main lobe position changes, and the steering angle gradually increases with the increase of phase difference. When the steering angle increases, the far-field main lobe energy gradually leaks into the grating lobes, which reduces the peak light intensity of the main lobe. When the peak intensity of the grating lobe is stronger than the main flap, the energy concentration of the steering beam on the far-field target surface is poor, which easily affects the position calculation of the far-field main lobe and interferes with the precise pointing control of the steering beam. Therefore, the limit range of steering is defined when the peak intensity ratio of the main lobe to the grating lobe is equal to 1. When the fiber laser phased array steers along the x- and y-axes respectively, there are obvious differences in the far-field spot shape and steering range, which is caused by the asymmetric structure of the fiber laser phased model. In this paper, the phased array models with apertures 19, 133 and 703 have equivalent diameters. As the number of sub-aperture increases, the aperture spacing decreases and the steering range increases. Therefore, the parameters of the phased array steering system can be designed according to the actual application scenario, and the aperture size and aperture number can be selected reasonably. By studying the steering characteristics of numerous sub-aperture and meter aperture fiber laser phased arrays, this paper enriches the beam wavefront control ability of fiber laser phased array technology, which can be used for precise tracking of ultra-long-distance targets and fast beam coverage in a certain range.Numerous sub-aperture fiber laser array is one of the emerging technologies to build high power, high beam quality and equivalent optical large aperture. Realizing the common phase and even the fast and flexible beam deflection of array laser beam based on the precise phase control is the key to the application of the current fiber laser phased array technology. In this paper, the optical phase-controlled steering technology is combined with the fiber laser coherent combining system, and the beam steering characteristics of the numerous sub-aperture fiber array laser coherent combining systems are studied. The beam steering is realized by changing the phase difference between the adjacent sub-aperture of the collimated laser array. The far-field steering beam pattern distribution characteristics of 19, 133 and 703 aperture fiber laser phased arrays are compared and analyzed, and the steering limit range is defined and calculated accordingly. The results provide a theoretical basis for the subsequent experimental research on the precise pointing control of fiber laser phased arrays under long-range transmission.

2) On the premise of taking an additional texture camera, marker points are added in the measured field, and the mapping relationship between the 3D point cloud and the 2D texture image can be obtained from the camera imaging model. To further get rid of the constraints of adding markers, an unconstrained free texture mapping method is proposed for objects with rich textures. The idea is to perform feature matching between the object images captured by the left and right cameras. According to the corresponding relationship between feature matching points and the 3D point cloud, the PnP problem is solved to obtain the pose relationship for the establishment of the mapping relationship between the 3D point cloud and the 2D texture image and finally realizes texture mapping. Experiments have proved the feasibility of these two methods. The research fruits of this paper could provide a simple and easy means of color 3D information acquirement for the fields of cultural relics digitization and reverse engineering.After obtaining three-dimensional (3D) point cloud data of a measured scene using the binocular structured light projection measurement system, three texture (grayscale and color) acquisition and mapping methods have been explored in this paper for different scenes. Under the condition of no additional color imaging equipment, two texture acquisition methods are proposed respectively. On the premise of using an additional texture camera, the method using free texture mapping by adding marker points is proposed. Then, to get rid of the dependence on marker points, the unconstrained free texture mapping method using the object’s feature information is presented. This paper proposes three kinds of practical and feasible solutions from 3D point cloud data to mapping texture for different applicable scenes, and the feasibility of the proposed methods is proved by experiments. The free texture mapping method based on a binocular structured light system provides a simple and easy means of color 3D information acquirement for the fields of cultural relics digitization and reverse engineering.

This paper designed a plasmonic refractive index sensor which consists of gold nano cones and a gold film with a SiO2 film as spacer-layer. The surface plasmon resonance modes in the composite structure are studied by using the Finite Difference Time Domain method. The composite structure can stimulate not only localized surface plasmon, but also propagating surface plasmon. The energy of the incident electromagnetic wave is partially coupled to the localized surface plasmon through a single gold nano cone, and partially coupled to the propagating surface plasmon through a grating of gold nano cone array. The reflection spectra of the composite structure are simulated in the refractive index range of 1.30 to 1.40. It is found that the resonance wavelength has a linear relationship with the refractive index of the analyte, and the reflectivity at the resonance is almost zero due to the strong resonance coupling between localized and propagating surface plasmon. In addition, the full width at half maximum of propagating surface plasmon resonance mode is very narrow when the geometric parameters of gold nano cone are optimized. The sensitivity and figure of merit reach 770 nm/RIU and 113 RIU?1 respectively, and it has good refractive index sensing performance. The designed composite structure is expected to be widely used in the field of biochemical detection.A surface plasmon resonance refractive index sensor based on the coupling structure of gold nano cones and a gold film with a SiO2 film as spacer-layer is designed. The surface plasmon resonance modes in the composite structure are studied by using the Finite Difference Time Domain method. The composite structure can stimulate not only localized surface plasmon, but also propagating surface plasmon. The energy of the incident electromagnetic wave is partially coupled to the localized surface plasmon through a single gold nano cone, and partially coupled to the propagating surface plasmon through a grating of gold nano cone array. The reflection spectra of the composite structure are simulated in the refractive index range of 1.30 to 1.40. It is found that the resonance wavelength has a linear relationship with the refractive index of the analyte, and the reflectivity at the resonance is almost zero due to the strong resonance coupling between localized and propagating surface plasmon. In addition, the full width at half maximum of propagating surface plasmon resonance mode is very narrow when the geometric parameters of gold nano cone are optimized. The sensitivity and figure of merit reach 770 nm/RIU and 113 RIU?1 respectively, and it has good refractive index sensing performance. The designed composite structure is expected to be widely used in the field of biochemical detection.

Overview: With the rapid development of aerospace, energy power, petrochemical, and other fields, nickel-based alloy sheet welding technology has become one of the key factors determining the performance of core components. The welding of nickel base alloy sheet is sensitive to the heat input, and it is easy to cause element segregation and brittle phase precipitation, which will reduce the weld performance and produce welding deformation. This paper introduces the research progress of laser welding technology of nickel base alloy sheet, and summarizes the evolution of weld microstructure, changes of mechanical properties and corrosion resistance, and the rules of welding deformation under two kinds of welding technologies including laser autogenous welding and laser welding with filler wire of nickel base alloy sheet. The research of the autogenous laser welding process focuses on Ni-Cr and Ni-Cr-Mo alloys. The grain morphology and element segregation are analyzed, including refining microstructure and inhibiting the formation of precipitates, by means of adjusting the process parameters, using ultrasonic vibration, and using a low-temperature cooling process. The microhardness of the two kinds of alloy welds is better than that of base metal, because of the finer grains in the welds. Tensile strength at room temperature can reach about 90% of the base metal, but high-temperature tensile performance is comparable to the base metal. Ni-Cr alloy welded joints show good high-temperature plasticity. The relatively lower tensile strength of the welded joints is relative to the worse morphology of the weld surfaces. The fatigue properties and corrosion resistance of the Ni-Cr-Mo alloy welds are comparable to those of the base metal. The research of laser welding of nickel-based alloy sheets with filler wire focuses on the Ni-Cr-Mo alloy, and the grain morphology, element segregation, and its regulation are still the focuses of the research. The microhardness and room temperature tensile strength of the welded joints with filler wire are better than those of the base metal. The better room temperature tensile strength of the welded joints benefits from both the finer weld grains and the occurrence of the reinforcement. Corrosion tests show that the welded joints have comparable corrosion resistance to the base metal. Welding deformation of nickel-based alloy sheets includes shrinkage deformation, deflection, and angular deformation. Compare with the traditional arc welding process, laser welding shows lower heat input, and thus, it leads to smaller deformation. At present, the research of welding deformation of nickel-based alloy sheet mainly concentrates on the prediction of deformation through the finite element method and reducing deformation through process parameters adjustment, restraint intensity control, and utilizing auxiliary processes. Future research should focus on the prediction of weld microstructure and the propose of various adaptive control strategies for microstructure, mechanical properties and corrosion resistance by combining with advanced algorithms. Besides, developing new types of intelligent welding processes is also an important part.With the rapid development of aerospace, energy power, petrochemical, and other fields, nickel-based alloy sheet welding technology has become one of the key factors determining the performance of core components. The welding of nickel base alloy sheet is sensitive to the heat input, and it is easy to cause element segregation and brittle phase precipitation, which will reduce the weld performance and produce welding deformation. This paper introduces the research progress of laser welding technology of nickel base alloy sheet, and summarizes the evolution of weld microstructure, changes of mechanical properties and corrosion resistance, and the rules of welding deformation under two kinds of welding technologies including laser autogenous welding and laser welding with filler wire of nickel base alloy sheet. It is proposed that the prediction of weld microstructure should be considered in the future research, which should combine with advanced algorithms to propose the adaptive control strategy of microstructure, mechanical properties and corrosion resistance, developing a new intelligent welding process.

Overview: With the development of remote sensing satellite technology and the expansion of the market in remote sensing images (RSIs), content-based remote sensing image retrieval (RSIR) plays an irreplaceable role in many fields, such as economic and social development, resource and environmental monitoring, and urban life management. However, there are complex content and rich background information in the high-resolution remote sensing images, whose features extracted by convolutional neural networks are difficult to effectively express the salient information of the RSIs. For this problem in high-resolution RSIR, a self-attention mechanism based on cascading pooling is proposed to enhance the feature expression of convolutional neural networks. Firstly, a cascade pooling self-attention module is designed. Cascade pooling uses max pooling based on a small region, and then adopts average pooling based on the max pooled feature map. Compared with traditional global pooling, cascade pooling combines the advantages of max pooling and average pooling, which not only pays attention to the salient information of the RSIs, but also retains crucial detailed information. The cascade pooling is employed in the self-attention module, which includes spatial self-attention and channel self-attention. The spatial self-attention combines self-attention and spatial attention based on location correlation, which enhances specific object regions of interest through spatial weights and weakens irrelevant background regions, to strengthen the ability of spatial feature description. The channel self-attention combines self-attention and content correlation-based channel attention, which assigns weights to different channels by linking contextual information. Each channel can be regarded as the response of one class of features, and more weights are assigned to the features with large contributions, thereby the ability to discriminate the salient features of the channel is enhanced. The cascade pooling self-attention module can learn crucial salient features of the RSIs based on the establishment of semantic dependencies. After that, the cascade pooled self-attention module is embedded into the convolutional neural networks to extract features and optimize features. Finally, in order to further increase the retrieval efficiency, supervised Hashing with kernels is applied to reduce the dimensionality of features, and then the obtained low-dimensional hash code is utilized in the RSIR. Experiments are conducted on the UC Merced, AID and NWPU-RESISC45 datasets, the mean average precisions reach 98.23%, 94.96% and 94.53% respectively. The results show that compared with the existing retrieval methods, the proposed method improves the retrieval accuracy effectively. Therefore, cascade pooling self-attention and supervised hashing with kernels optimize features from two aspects of network structure and feature compression respectively, which enhances the feature representation and improves retrieval performance.In high-resolution remote sensing image retrieval, due to the complex image content and rich detailed information, it is difficult for the features extracted by a convolutional neural network to effectively express the salient information of the image. In response to this issue, a self-attention module based on cascade pooling is proposed to improve the feature representation of convolutional neural networks. Firstly, a cascade pooling self-attention module is designed, and the self-attention module can learn key salient features of images on the basis of establishing semantic dependencies. Cascade pooling uses max pooling based on a small region, and then adopts average pooling based on the max pooled feature map. The cascade pooling is exploited in the self-attention module, which can keep important details of the image while paying attention to the salient information of the image, thereby enhancing feature discrimination. After that, the cascade pooled self-attention module is embedded into the convolutional neural network for feature optimization and extraction. Finally, in order to further improve the retrieval efficiency, supervised hashing with kernels is applied to reduce the dimensionality of features, and then the obtained low-dimensional hash code is utilized for remote sensing image retrieval. The experimental results on the UC Merced, AID and NWPU-RESISC45 data sets show that the proposed method can improve the retrieval performance effectively.

The Lukosz pre-correction modal algorithm can correct low-order aberrations of wavefront distortion and narrow the search range of iterative algorithms. The adaptive cosine-decay stochastic parallel gradient descent (AcSPGD) algorithm can compensate for high-order aberrations of wavefront distortion and improve the correction accuracy of iterative algorithms. In this paper, a new hybrid modal algorithm based on the pre-correction model and AcSPGD algorithm is applied to correct wavefront distortion in wavefront sensorless adaptive optics, and the feasibility of the optimization algorithm is also verified by the experiments. Experimental results show that the correction speed of the hybrid modal algorithm is two times faster than the commonly used stochastic parallel gradient descent (SPGD) algorithm, and the correction accuracy of the hybrid modal algorithm is better than the traditional Lukosz modal algorithm. Applied to wavefront sensorless adaptive optics, the optimization algorithm effectively reduces the phase fluctuation of the wavefront and improves the far-field Strehl ratio, which thus improves the signal-to-noise ratio of the atmospheric laser communication system by 2.9 dB, reduces the bit error rate to 10?6, and improves the communication performance of the free-space optical communication system. The hybrid modal algorithm has great reference and application value.The Lukosz pre-correction modal algorithm can correct low-order aberrations of wavefront distortion and narrow the search range of iterative algorithms. The adaptive cosine-decay stochastic parallel gradient descent (AcSPGD) algorithm can compensate for high-order aberrations of wavefront distortion and improve the correction accuracy of iterative algorithms. In this paper, a new hybrid modal algorithm based on the pre-correction model and AcSPGD algorithm is applied to correct wavefront distortion in wavefront sensorless adaptive optics, and the feasibility of the optimization algorithm is also verified by the experiments. Experimental results show that the correction speed of the hybrid modal algorithm is two times faster than the commonly used stochastic parallel gradient descent (SPGD) algorithm, and the correction accuracy of the hybrid modal algorithm is better than the traditional Lukosz modal algorithm. Applied to wavefront sensorless adaptive optics, the optimization algorithm effectively reduces the phase fluctuation of the wavefront and improves the far-field Strehl ratio (SR), thus improving the communication performance of the free-space optical communication (FSO) system.

Overview: Quantum dot is a kind of semiconductor nanocrystal with a quantum confinement effect. Recently, quantum dots have been applied for display due to their advantages including high photoluminescent efficiency, tunable emission wavelength, narrow emission spectrum, and low-cost solution process. In this paper, we focus on the application of quantum dots in microdisplay. With the rising near-eye display demands such as AR/VR, the realization of full color, high efficiency, and high luminance microdisplay attracts many attentions. However, the realization of the target microdisplay is difficult due to the high-cost mass transfer technology in micro-LED and the insufficient luminance in micro-OLED. Here, the photoluminescent (PL) and electroluminescent (EL) quantum dots can provide new routes for microdisplay. For PL, quantum dots can work as color conversion material for micro-LED. The multiple-time mass transfer can be avoided with the combination of red and green quantum dots and blue micro-LED. Meanwhile, the color gamut can be improved due to the narrow FWHM of quantum dot emission. For EL, RGB quantum dots can work as an emission layer in QLED, and the RGB micro-QLED can be applied for microdisplay directly with a compact and high-efficiency system. Compared with OLED, the QLED can realize higher luminance due to the inherent stability of the inorganic quantum dot, which is more suitable for AR display requiring high luminance. The patterning of quantum dot layer is the first step for the application in microdisplay. For both PL and EL applications, a high pixel density, high pixel uniformity, high pixel consistency, and low-cost patterning method is required. For the quantum dot color conversion layer in PL application, a high optical density is required for the sufficient absorption of the blue light. For the quantum dot emission layer in EL application, the uniform and small roughness surface quantum dot layer is required with few damages to the quantum dots to ensure the good performance of QLED devices. There are several patterning methods have been reported for quantum dots including inkjet printing, photolithography, transfer printing, electrophoretic deposition, in situ fabrication, and optical micro cavity. However, it is still challenging to find a perfect patterning method for the quantum dot layer. For PL application, the stability of quantum dot under long time high-intensity blue light excitation is a big problem due to the photoinduced quenching and oxidation. For EL application, compared with red and green QLED, the peak luminance, efficiency, and lifetime of blue QLED needs to be further improved by optimizing the blue quantum dots and device structure to satisfy the requirement of the display application.The quantum dot is a kind of semiconductor nanocrystal with quantum confinement effect, which attracts a lot of attention due to the excellent optoelectronic properties and has been widely used in the display area. The quantum dot become one of the core materials of display with several advantages including the high luminance efficiency, tunable emission wavelength, narrow FWHM and low-cost solution fabrication. Micro display technology is applied in the near eye display scenario with small effective display area (diagonal < 1 inch). Recently, the rising of VR/AR application scenarios require micro display technology with higher luminance, higher pixel density, and full color display. In this paper, we review the current progress of the quantum dot in micro display from photoluminescence and electroluminescence technique routes. The chances and challenges of the quantum dot in micro display are also summarized.

Overview: Optical super-resolution lenses have shown great potential in super-resolution microscopic systems and nano-fabrication systems. With the decrease of the focusing spot of the super-resolution lens, it is inevitable that large sidelobes and sidebands will be generated, which will lead to a limited field of view and imaging artifacts. Therefore, when designing super-resolution optical devices, it is necessary to adopt a balanced strategy between focusing spot and side lobe according to the practical applications. Metasurface is a planar structure composed of nanoscale meta-atoms, which can flexibly regulate the amplitude, phase and polarization of the optical field, being beneficial to construct complex super-resolution optical fields. The PB phase meta-atom is comparatively easy to fabricate due to its simplicity. Using Finite-Difference Time-Domain (FDTD) solutions to optimize the size of the meta-atom, we can get a structure with high transmittance. By rotating the angle of the meta-atom, we can achieve linear phase control. The application of PB phase metasurface has been demonstrated in the field of super-resolution focusing devices with suppressed sidelobe. Based on the vector angular spectrum method and particle swarm optimization (PSO) algorithm, a super-resolution point focusing lens with a large numerical aperture and weak sidelobe is optimally designed with a 32-valued phase control at the wavelength of λ=632.8 nm. Based on the silicon-based PB phase metasurface, our metalens was fabricated by electron beam lithography and orthoplastic etching. The lens radius Rlens=57λ, focal length zf=20λ, corresponding to the numerical aperture of NA=0.944. The optical field distribution of the super-resolution metalens was measured experimentally by a large-numerical-aperture microscopy system. The results show that, at the focal plane, the FWHM of the focal spot is 0.45λ, which is less than the diffraction limit of 0.53λ (the diffraction limit is 0.5λ/NA), the side-lobe ratio SR is 0.07, and the depth of focus is 0.4λ. Our proposed metalens can achieve a small depth of focus, a weak sidelobe ratio, and super-resolution point focusing. Our proposed super-resolution metalens bears the potential to realize the miniaturization, lightweight, and integration of super-resolution optical devices or systems.Metasurface is a spatially varying ultrathin nanostructure that has been widely studied and used in optical super-resolution focusing, either in lenses or in systems. However, with the decrease of the focal spot size of the metalens, large sidelobes are inevitably generated, limiting the field of view and potential applications of the lens. In this paper, a method for producing super-resolution metalens with a large numerical aperture (NA=0.944) and weak sidelobe is presented. For a circularly polarized light with the wavelength of λ=632.8 nm, a super-resolution point-focusing with a weak sidelobe is realized based on PB phase regulation of silica-based metasurface. Experimental results show that the FWHM (full-width at half maximum) of our focusing spot is 0.45λ, which is less than the diffraction limit of 0.53λ (the diffraction limit is 0.5λ/NA), and the sidelobe ratio (SR) is 0.07. Our proposed super-resolution metalens bears the potential to realize the miniaturization, lightweight and integration of super-resolution optical devices or systems.

Overview: Surface plasmon resonance (SPR) sensing technology has attracted widespread attention due to its advantages of high sensitivity, label-free, and real-time dynamic monitoring. Traditional SPR sensing platform needs the use of a prism, and requires that the transverse magnetic (TM) polarized light incident at a specific angle to satisfy the wave vector matching condition and excite the surface plasmon polariton (SPP) mode at the interface between the metal film and the external environment. Moreover, Tamm plasmon polariton (TPP), as a special plasmon boundary state mode, can be excited by using the boundary between the one-dimensional Bragg photonic crystal (PC) and the metal film and has broad application prospects in the fields of new optoelectronic devices. Compared with SPP, the excitation of TPP does not require wavevector compensation for incident light and can be achieved at any polarization. However, the enhanced electromagnetic field of the TPP mode is mainly localized inside the structure and cannot sense the changes in the external environment, which greatly limits its application in the field of biochemical sensing. To break through this limitation, researchers integrated the one-dimensional Bragg PC structures onto the traditional prism structures to achieve hybrid coupling of SPP mode and TPP mode by using the oblique incident light, which could improve the sensing performance of the SPR sensors. However, this kind of TPP-SPP strong coupling excitation also requires a bulky prism and a precise incident light angle control system, which is not conducive to the miniaturization and integrated application of the structure. Therefore, we propose a feasible design of a grating-coupled multilayer stack in this paper. The structure mainly consists of three parts: a nanometric gold film on the top layer, a one-dimensional Bragg PC in the middle, and a gold nanograting on the bottom. In this structure, the SPP and TPP resonance excitations on the upper and lower surfaces of the top nano-gold film are simultaneously achieved by utilizing the first-order transmitted light of the bottom nanograting. The coupling hybridization between the two modes greatly reduces the resonance bandwidth of the generated hybrid mode, resulting in a significant improvement in its sensing figure of merit. In addition, the coupling hybridization of the SPP and the TPP can be realized in a wide spectral range by changing the period of the nanograting and the thickness of the dielectric layers constituting the one-dimensional Bragg PC. Compared with the traditional prism TPP and SPP dual-mode coupling structure, the designed multilayer nanostructure can realize the resonance coupling of the two modes over broad wavelength ranges at the normal incidence. These results not only make it easier to further integrate and miniaturize the structure, but also have important significance for broadening the practical application of the surface plasmon resonance sensors.The hybrid coupling of Tamm plasmon polariton (TPP) and surface plasmon polariton (SPP) on the surface of a gold film based on the prism coupling has attracted extensive attention and has been widely investigated. However, the traditional excitation configuration has bulky optical elements and requires accurate control of the angle of incident light, which limits its integration and practical application. In order to simplify the excitation condition of the TPP-SPP hybrid mode, a feasible grating-coupled multilayer stack structure is proposed in this paper. The structure mainly consists of three parts: a nanometric thin gold film on the top layer, a one-dimensional Bragg photonic crystal in the middle, and a gold nanograting on the bottom. In this structure, the SPR and TPP resonance excitations on the upper and lower surfaces of the top gold film are simultaneously achieved by utilizing the first-order transmitted light of the bottom nanograting. The hybrid coupling between the two modes greatly reduces the resonance bandwidth of the generated mode, thereby significantly improving the sensing figure of merit of the generated mode. Additionally, the hybrid coupling of both SPP and TPP modes can be realized in a wide spectral range by altering the period of the nanograting and the thickness of the one-dimensional Bragg photonic crystal. Compared with the traditional prism-coupled TPP and SPP dual-mode coupling structures, the designed grating-coupled multilayer nanostructure can realize the resonant coupling of the two modes at the normal incidence without prism and limitation of incident angle. This not only facilitates the further integration and miniaturization of the structure, but also has important significance for broadening the practical application of surface plasmon resonance sensors.

Overview: Optical coherence, as a fundamental resource in all areas of optical physics, plays a vital role in understanding interference, propagation, scattering, imaging, light-matter interactions, and other fundamental characteristics from classical to quantum optical wave fields. The theory of optical coherence is the most powerful tool to describe the statistical characters of random light beams (also named partially coherent beams). In the space-frequency domain, the spatial coherence property of a partially coherent light beam is characterized by a two-point spectral degree of coherence that is a normalized version of the cross-spectral density function. Nowadays, the degree of coherence has been viewed as a novel degree of freedom for the structured partially coherent light beams, which is akin to the deterministic properties, such as the amplitude, phase, and polarization of a fully coherent structured light beam. Due to the fundamental difference between the two-point degree of coherence of partially coherent light and the one-point deterministic features of fully coherent light, the partially coherent beams with customized spatial coherence have shown many unique properties and been found to be more advantageous in particular applications. By simply adjusting the spatial coherence width of the degree of coherence for a partially coherent beam can help reduce the turbulence-induced signal distortion in free-space optical communications and resist the speckle noise in optical imaging. Only recently, it has been found that not only the spatial coherence width but also the spatial coherence distribution of the degree of coherence can be customized, which has enabled a host of novel physical effects, including beam’s self-shaping, self-reconstruction, and self-focusing, and has aroused many important potential applications. In this paper, we review the fundamental theory and efficient experimental protocols for tailoring the spatial coherence structure of the degree of coherence for the partially coherent light beams. The differences and the advantages between the two strategies for producing the partially coherent beams with nonconventional spatial coherence structures are discussed. Meanwhile, we mainly focus on the applications of the spatial coherence structure engineering in coherence-based optical encryption, robust optical imaging, sub-Rayleigh imaging, robust far-field information transfer, and high-quality beam shaping. It is found that the spatial coherence structure engineering provides an efficient degree of freedom for the manipulation of structured light and paves the way for resisting the side effects induced by random fluctuations of complex media. We prospect that the spatial coherence engineering protocols can be extended to the temporal domain or even to the spatiotemporal domain and will find broader applications for light manipulations and light-matter interactions.Structured light has rich adjustable spatial degrees of freedom, including amplitude, phase, polarization, degree of coherence, etc. The modulation of these degrees of freedom has triggered a variety of novel physical effects and has found use in constructing new structured light beams and a large range of applications. Compared to the fully coherent light, partially coherent beams (PCBs) have advantages in resisting the speckle noise and the fluctuations of atmospheric turbulence. Recently, the PCBs with nonconventional coherence structures have been found to have important potential applications in atmospheric transmission, optical encryption and imaging, robust information transmission, and high-quality beam shaping. In this review, we summarize in detail the progress of the theoretical construction and experimental generation of PCBs with novel coherence structures. Meanwhile, we outline their robust propagation properties in complex media and important applications in optical encryption, imaging, robust information transfer, and beam shaping. It is found the modulation of spatial coherence structure of PCBs provides not only an efficient way to resist the random fluctuations of complex environments, but also a new degree of freedom to enrich the application scopes of structured light. Finally, the development trend and the further applications of the nonconventional coherence structure engineering are prospected.

Overview: Polarization holography has important application prospects in the field of data storage and polarized light imaging due to its ability to record amplitude, phase, and polarization information. In addition, it also has the ability to regulate light fields, which can regulate special light fields with helical phase distribution and spatial polarization distribution. Such special light fields have broad application prospects in the fields of optical communication, particle manipulation, photon entanglement, etc. There is also a lot of researches focused on how to generate such beams, such as helical phase plates, mode conversion, spatial light modulators, etc. However, the traditional method requires the construction of a relatively large optical system, which limits its application in fields such as integrated optics. The introduction of the beam preparation method of polarization holography can reduce the volume of the optical system to a certain extent. At the same time, the use of polarization-sensitive materials with the ability to record multi-dimensional information greatly reduces the cost on the one hand. On the other hand, it is easy to operate during the preparation process, which is expected to be an ideal material for beam preparation to some extent. Based on the introduction of the principle of faithful reconstruction of any polarization state by polarization holography, this paper reviews the research progress of generating vector beams, scalar vortex beams, and vector vortex beams based on polarization holography in the past two years. Faithful reconstruction for any polarization state refers to under the incident into the polarization-sensitive material at 90 degrees interference angle between the signal and reference waves, the recording and reading waves are p-polarized and the reconstruction wave can be reconstructed correctly. Phenanthrenequinone-doped polymethyl methacrylate photopolymer (PQ/PMMA) is used as a recording material in the experiment. First, the single control ability of polarization holography in polarization and phase is demonstrated respectively, and then the ability of polarization holography to control both polarization and phase at the same time is further introduced. Based on the characteristics of polarization holography, the signal optical path is regulated, and the vector beam, scalar vortex beam, and vector vortex beam are generated by setting the initial azimuth angle of the rotating components and adjusting their relative rotational angular velocity under dynamic exposure. In the fabrication process, the desired beam can be generated by simply controlling the parameters of some devices. Finally, the ability and prospect of generating special light fields based on polarization holography are briefly summarized and discussed.Polarization holography has important application prospects in the field of data storage and polarized light imaging due to its ability to record amplitude, phase and polarization information. In addition, it also has the ability to regulate light fields, which can regulate special light fields with helical phase distribution and spatial polarization distribution. Such special light fields have broad application prospects in the fields of optical communication, particle manipulation, photon entanglement, etc. There is also a lot of research focused on how to generate such beams. The latest research progress in preparing vector beams, scalar vortex beams, and vector vortex beams by using polarization holography is introduced in this paper. The light field regulation method based on polarization holography has the advantages of a simple fabrication process, the small size of the optical system and low production cost, which provides a new idea for the manufacture of special light fields.

Overview: Featured by the capability of multi degree-of-freedom light-field manipulations while reserving high spatial resolution, multifocal laser arrays have been widely applied in femtosecond laser micro/nanofabrication, optical trapping, etc. However, for lens diffraction, the smaller momentum spread along the optical axis with respect to that in the transverse direction could introduce a larger position spread in real space, which in turn leads to lower axial resolution than the transverse resolution. The anisotropy of the focused laser beam, inherent regardless of paraxial or tight-focusing cases, has been a great hurdle for laser printing of functional microdevices with precise control on feature size and improved mechanical performances. To this end, in this research, a feasible method for generation of isotropic focused laser beam with quasi-spherical 3D point spread function (PSF) is developed based on vectorial light field modulation. We demonstrate that through simultaneous implementation of phase modulation and amplitude modulation, homogeneous multifocal array with quasi-spherical focal spots can be generated. Particularly, with the use of a well-designed annular mask, the suppression on the axial spread of field is accomplished via accurate control on the coherent superposition of the orthogonal radially polarized beam (RPB) and azimuthally polarized beam (APB) in the focal region since the depolarized axial component of the AP beam vanishes in vicinity of the gaussian focus even under tight focusing condition. Using the proposed method, isotropic 3D PSF with identical axial and transverse FWHM of 0.71λ is achieved. Meanwhile, based on iterative phase retrieval algorithm, phase-only holograms are designed and employed transforming the incident wavelet as the summation of sub-wavelets, yielding multiple converging sites in 3D space, thereby generating the multifocal array. We further present the synthesis of quasi-spherical multifocal array. A high uniformity up to 99% for a 10-by-10 multifocal array, in which the single focus elements share near-identical axial and transverse FWHM, being 0.76λ on average. The standard deviation of the axial and transverse FWHM of the multifocal array are evaluated be 0.005λ and 0.019λ, respectively, highlighting the features of high uniformity and isotropy. The reported strategy renders precise control on the axial feature size and is potential for the application in high-precision parallel laser printing technique.Featured by the capability of multi degree-of-freedom light-field manipulations while reserving high spatial resolution, multifocal laser arrays have been widely applied in femtosecond laser micro/nanofabrication, optical trapping, and so forth. Yet, due to the relatively lower axial resolution of single focuses within the array in comparison with the lateral resolution of their own, multifocal laser array has been refrained from isotropic 3D nanofabrication. Herein, we propose a feasible method for generation of axially super-resolved multifocal array with quasi-spherical focal spots. In particular, quasi-spherical multifocal array is optically synthesized via precise modulation on the coherent superposition of the orthogonal radially polarized beam (RPB) and azimuthally polarized beam (APB) states in the focal region based on annular amplitude modulation. We show theoretically the generation of quasi-spherical multifocal array with a high uniformity up to 99%. The average axial and lateral full-width-half maximum (FWHM) of the focal array are measured to be 0.76λ with the standard deviations in the axial and lateral directions being 0.005λ and 0.019λ, respectively. The presented strategy for synthesis of quasi-spherical multifocal array with high uniformity paves the way for high-precision laser fabrication of 3D micro/nano devices.

Overview: In recent years, vortex beams have become the focus of research, and their orbital angular momentum makes them have many important applications, like optical communication, particle manipulation, and optical measurement. At the same time, researchers are paying attention to more abundant generation methods. In previous studies, vortex beam generation methods are usually divided into two categories. The first category is the outcavity, such as spiral phase plate method, spatial light modulator method, mode conversion method, metasurface method, and corner array method, and the second category is the incavity, such as point-loss method, off-axis pumping method, and spatial light modulator method. However, these methods can not tolerate high power laser output and adjust topological charges flexibly. Therefore, how to generate a vortex beam that can tolerate high power laser output and adjust the topological charges flexibly is an important problem to be solved. Continuous surface deformation mirror is a key component of adaptive optical system. In the study of wavefront fitting for continuous surface deformation mirrors, there are usually two kinds of methods. The first type is model-free method, such as genetic algorithm, simulated annealing algorithm, stochastic parallel gradient descent (SPGD) algorithm, etc. These methods generally require many iterations and slow convergence, and it is difficult to change the topological charge flexibly. The second type is pattern method, such as Zernike mode method, Lukosz mode method, and enginmode method. This method first defines a set of complete orthogonal modes, calculates the mode coefficients, and completes the fitting of the target wavefront by linear superposition of each mode. Zernike mode is orthogonal in the circular domain, Lukosz mode is orthogonal in the circular domain derivative. However, usually the configuretion of deformation mirror is not circular domain. For example, the deformation mirror driver used in this paper is arranged in circular domain. In this case, the orthogonal basis needs to be rebuilt to use these two methods. The eigenmode of the deformed mirror is directly and precisely derived from the influence function of the deformed mirror drivers, so it can not only avoid the influence of fitting error, improve the fitting accuracy, but also adapt to the different configuration of the deformed mirror. Combined with the eigenmode method, continuous surface deformation mirror can fit all kinds of vortex beams with high precision and fast fitting speed, and can be applied to all kinds of deformation mirrors with different configurations. In this paper, the eigenmode method of continuous surface deformation mirror is used to simulate and analyze the fitting of the spiral wavefront of integer order with topological charge is ?5 to 5, fractional order, multi-fractional order, and superposition state with the absolute value of topological charge less than 5. Various vortex light fields are generated by dynamic manipulation. The results show that the continuous surface deformation mirror will have a good application prospect in the field of high-power vortex field manipulation.A complete orthogonal basis was constructed by using the eigen-mode method of continuous surface deformation mirror, and the voltage of each driver of the deformation mirror can be obtained according to the spiral wavefront information which needs to be manipulated. The spiral wavefront of integral order, fractional order, multi-fractional order, and superposition state with the absolute value of topological charge less than 5 was generated, and the dynamic manipulation of the vortex beam was realized. The results obtained were the same as those obtained by the ideal spiral wavefront. The ability of the continuous surface deformation mirror to fit the spiral wavefront was demonstrated and good results were obtained. This method has a good application prospect in the dynamic manipulation of high-power vortex laser.

Overview: The spatial resolution of traditional optical microscopy is limited by the diffraction limit λ/(2NA) (λ is the wavelength, NA is the numerical aperture of the objective lens of system), and the lateral resolution is about 200 nm~300 nm, which makes it difficult to achieve clear imaging for micro-nano structures or cell samples. In this paper, a label-free far-field super-resolution imaging method based on hyperbolic metamaterial is proposed. Super-resolution optical microscopy is an important technology due to the non-contact and non-destructive advantages. Currently, most of the super-resolution imaging methods rely on the fluorescent dyes, which limited their applications. The label-free far-field microscopy imaging method based on the frequency shift effect has been proposed and developed in recent years. However, its spatial resolution is limited by the refractive index of waveguide materials. Based on the characteristic of optical spatial spectrum band-pass filtering in hyperbolic metamaterials (HMM), a large-area uniform bulk plasmon polariton (BPP) field with high spatial frequency can be achieved by combining with nano-scale gratings. Due to the large wave vector of the BPP illumination, the high-frequency information of the object can be transferred to the passband in traditional imaging systems and participate in super-resolution imaging. Illuminated by a BPP field with 2.66k0 at the wavelength of 532 nm, a double-slits structure with a 100 nm-wide center-to-center distance has been resolved with a 0.85 numerical aperture standard objective based on this method. The lateral resolution is improved to λ/5.32. By further improving the transverse wave vector of BPP, it can be improved to λ/7.82. This design is label-free and conveniently integrated with traditional microscopes, which provides a visual super-resolution imaging method for applications in biomedicine, on-chip industry, material science, and other fields.Super-resolution optical microscopy is an important technology due to the non-contact and non-destructive advantages. Currently, most of the super-resolution imaging methods rely on fluorescent dyes, which limited their applications. The label-free far-field microscopy imaging method based on the frequency shift effect has been proposed and developed in recent years. However, its spatial resolution is limited by the refractive index of waveguide materials. Based on the characteristic of optical spatial spectrum band-pass filtering in hyperbolic metamaterials (HMM), a large-area uniform bulk plasmon polariton (BPP) field with high spatial frequency can be achieved by combining with nano-scale gratings. Due to the large wave vector of the BPP illumination, the high-frequency information of the object can be transferred to the passband in traditional imaging systems and participate in super-resolution imaging. Illuminated by a BPP field with 2.66 k0 at a wavelength of 532 nm, a double-slit structure with a 100 nm-wide center-to-center distance has been resolved with a 0.85 numerical aperture standard objective based on this method. The lateral resolution is improved to λ/5.32. By further improving the transverse wave vector of BPP, it can be improved toλ/7.82. This design is label-free and conveniently integrated with traditional microscopes, which provides a visual super-resolution imaging method for applications in biomedicine, on-chip industry, material science, and other fields.

Overview: As an ideal 3D display technology, holography can reconstruct the wavefront of the whole light wave, and can provide all the 3D depth cues required by the human eyes, including binocular parallax, motion parallax, accommodation, occlusion, etc. Due to the limitation of the modulation principle, DMD and most SLM cannot optically reconstruct the complex amplitude of a wavefield, resulting in partial information loss and complex wavefront calculation. At the same time, the two devices have a pixel size larger than 6 μm, which is much larger than the wavelength of visible light. The limitation of large pixel size and modulation principle brings many disadvantages, such as narrow field of view, twin-image, narrow band, and multi-order diffraction, which greatly restrict the development of CGH. As a new class of light field modulators, metasurface can control the amplitude, phase, polarization and dispersion of the light simultaneously by optimizing the design and arrangement of the elements. Thanks to the previous exploration of micro-nano manufacturing technology and materials for metasurface, the size of the unit cell can be reduced to the order of sub-wavelength. According to the grating equation, the smaller the pixel size is, the larger the diffraction angle is. Therefore, metasurface can provide a diffraction angle close to 90°. As the loading medium of holograms, metasurface meets the requirements of holograms for high-precision and complex light field modulation and has the advantages of high design freedom, high spatial resolution, low noise, broadband and so on, providing a solution to some problems currently faced by CGH. In this paper, the basic process of designing meta-holography devices is discussed. Furthermore, the basic concepts and development of static meta-holography are introduced based on the principles of metasurfaces, including phase modulation, amplitude modulation, complex-amplitude modulation, and nonlinear modulation. However, such static meta-holography devices cannot change the display patterns after design and manufacture, which is inconsistent with the rapidly changing real world and requirements of diverse functions, limiting its applications. Therefore, the two methods of realizing dynamic meta-holography are introduced in detail. Finally, the micro-nano fabrication technologies for metasurface are discussed. In conclusion, this paper presents the design, principle, development, and manufacturing implementation of meta-holographic devices in an all-around way, and puts forward problems and possible solutions for the development of meta-holography at present.The pursuit of real-time, full-color, three-dimension (3D), and dynamic display has inspired a rich body of industrial and academic research. With the introduction of "Metaverse", there is an increasing demand for high-performance 3D display devices and technologies. Holographic technology is an ideal approach for future naked-eye 3D display. However, traditional dynamic holographic devices have brought many shortcomings such as small field of view (FOV) and limited information capacity, which hinder the practical applications. As a new class of light field modulator, metasurface is expected to achieve remarkable breakthroughs in the field of holographic display with the advantages of their small pixel size and the emerging ability to manipulate light. This paper gives an overview of the development of meta-holography from four aspects: the design strategy, the modulation principle, the methods for realizing dynamic display and the micro-nano fabrication technologies for optical metasurface. We finally include a brief discussion of the future direction in this field.

Overview: Photonic integrated circuits (PIC) serve as an essential and promising candidate to eventually replace electronic circuits for the next-generation information processing. However, traditional PIC devices based on optical waveguides are usually bulky and lack full control at the subwavelength scale to achieve arbitrary wavefront-shaping functionalities. Recently, the invention of on-chip metasurface promotes the connection between guided and free-space optics and realizes the arbitrary conversion of guided waves and free-space light. As a new type of on-chip nanophotonic device, the introduction of metasurface onto the optical waveguide has made significant progress and exhibited multi-functional conversion from the guided waves to free-space, including directional beam-steering emitters, mode-conversion, on-chip lensing, optical router, and on-chip holography, etc. These on-chip nanophotonics devices provide new avenues for photonic chip-scale devices and miniature on-chip systems. For instance, meta-holography is an emerging and universal strategy based on engineered nanoantennas array to construct an optical-field image. However, on-chip far-field holographs are limited for realizing multiplexing for multiple directions due to a lack the arbitrary-encoding capability because their detour phases are complementarily related when the source propagates and excites the on-chip array from either positive or negative direction. Here, we propose and experimentally demonstrate a quad-fold multiplexed far-filed holographic display optics device based on an on-chip metasurface. This optics device is composed of silicon nanopillar arrays on top of a planar waveguide of Si3N4, in which a relatively thick layer of silica serves as the bottom cladding substrate. By mixing the detour phase and Pancharatnam-Berry (PB) phase, the on-chip metasurface could couple the guided waves into free space in circular polarization. The phase degeneracy in the positive and negative directions could be decoupled by selecting the desired circular polarization. Subsequently, utilizing a simulated annealing phase optimization algorithm to optimize the phase required by holograms and the multiplexing technology of on-chip directional, we achieved a quad-fold multiplexed far-field holographic display with independent encoding capability. Eventually, to verify the on-chip quad-fold multiplexed holography performance, we fabricated an on-chip metasurface sample by the conventional electron-beam lithography technique and the reactive ion etching processing. Through end-fire coupling from the laser source at λ = 650 nm into the on-chip metasurface sample along ±x/±y - directions, the far-field holographic images of the four letters (“A”, “B”, “C”, and “D”) multiplexing are successfully observed at their corresponding areas. The method proposed here opens up new prospects for the multifunctional integration of on-chip metasurfaces and provides an alternative approach for integrated optical communication with high information storage capacity.The on-chip metasurface is introduced into integrated optical waveguides to achieve arbitrary modulation of guided waves, which provides a convenient and versatile platform for the conversion between guided waves and free-space functions. Despite previous explorations in on-chip holography demonstration, it still faces critical challenges to expand the encoding freedom and multiplexing. Here, we propose and experimentally demonstrate a quad-fold multiplexed holographic display optics device based on an on-chip metasurface. By mixing the detour phase and Pancharatnam-Berry (PB) phase, the on-chip metasurface couples the guided waves into free space in circular polarization, destroying the phase degeneracy that exists in the wavevector directions with only the detour phase. Moreover, by utilizing simulated annealing phase optimization algorithm and multiplexing, we achieved a quad-fold multiplexed far-field holographic display with independent encoding capability. The proposed method in this paper opens up a new prospect for multifunctional integration of on-chip metasurfaces and provides an alternative approach for integrated optical communication with high information storage capacity.

Overview: Image edge extraction is a widely used and rapidly developing technology, playing an important role in medical imaging, enhanced vision, automatic driving and other fields. In recent years, there has been growing interest in developing miniature metasurface devices to obtain image edge information. Currently, it has been reported that discrete metasurface edge detection devices are used to obtain image edge information, but discrete metasurfaces often maintain a high energy efficiency only near the preset wavelength, and the energy efficiency decreases when deviating from the preset wavelength, which will limit the operating bandwidth of the metasurface optical computing device. Here, an optical differential device is designed by using a metasurface composed of quasi-continuous nanostrips to realize one-dimensional images edge detection. By changing the spatial orientation of quasi-continuous nanostrips, the device achieves geometric phase in the range of 0~2π, and maintains high energy efficiency over a wide wavelength range. The optical path system consists of two linear polarizers and two lenses with the same focal length, of which two lenses are placed in a confocal position to form a classical 4f optical system. The designed quasi-continuous metasurface edge detection device is placed on the Fourier plane of the 4f optical system. The original image is located on the object plane of the 4f optical system (at the front focal plane of the lens 1), and the object edge information is finally obtained on the image plane of the 4foptical system (at the rear focal plane of the lens 2). The simulation results show that the designed sample can achieve high average energy efficiency edge detection in the whole visible and near-infrared bands. Specifically, the quasi-continuous meta-device can obtain a clear image of object edge in the wavelength range of 400 nm~1000 nm, the energy efficiency of the device reaches 90.27% at the wavelength of 600 nm, and the average energy efficiency is 64.57% at the wavelength of 400 nm~1000 nm. Compared with the traditional edge detection devices based on discrete metasurface, the quasi-continuous devices have higher broadband average energy efficiency. Hopefully, this work enjoys many research merits in signal processing, optical communication and machine vision.In this paper, we design an optical differential device based on quasi-continuous metasurface and realize one-dimensional edge detection of an optical image. By changing the spatial orientation of quasi-continuous nanostrips, the device achieves geometric phase in the range of 0~2π, and maintains high energy efficiency over a wide wavelength range. The simulation results show that when the illumination wavelength increases from 400 nm to 1000 nm, the quasi-continuous meta-device can achieve clear images for the target edge. The maximum energy efficiency is 90.27% (the incident wavelength is 600 nm) and the average energy efficiency is 64.57% (the incident wavelength changes from 400 nm to 1000 nm). It can be expected that the proposed method can promote the application of quasi-continuous metasurface in image information processing and ultrafast optical computation.

Overview: Metasurface is a new kind of artificial two-dimensional material. Its working principle is to flexibly control the amplitude, phase and polarization of the incident electromagnetic wave by using the local interaction between the subwavelength scale unit cell and electromagnetic wave. Compared with traditional optical devices, devices based on metasurfaces have the advantages of compact structure, diverse functions, and easy integration. Therefore, metasurface has become a research hotspot in optics and photonics. At present, the electromagnetic manipulation devices based on the metasurfaces have achieved many novel functionalities, such as perfect absorption, anomalous deflection, focused imaging, electromagnetic cloak, and high efficiency holography. However, there are still some key problems to be solved in this field such as low working efficiency and narrow bandwidth. In recent years, the emergence of catenary electromagnetics provides new ideas and methods to solve these problems. In fact, catenary was first used in engineering and architecture to describe the shape of a soft rope suspended under the uniform gravitational force between two horizontal points. The use of catenary equations to solve problems in the field of electromagnetism has only recently been discovered by researchers. In this paper, we proposed a metasurface absorber based on a twisted catenary structure in the near-infrared band. The local electric field enhancement effect of the structure is different when the incident electromagnetic wave is with opposite spins, which can achieve efficient chiral selective absorption. The simulation results show that the circular dichroism is larger than 78% at the working wavelength. At the same time, the designed structure also has good angular stability, and can still get larger than 60% circular dichroism absorption in the case of oblique incidence at different azimuth angles. Besides, a possible method of information encryption using this kind of structure is proposed. Different information can be read by controlling the handedness of incident electromagnetic wave. This work further enriches the content of catenary electromagnetics, and has certain research value in the fields of chiral imaging and chiral sensing.As a kind of artificial two-dimensional material, metasurfaces have drawn wide attentions in recent years due to their ultra-thin profile and flexible electromagnetic manipulation capability. Therefore, how to further improve the working efficiency of metasurface devices has become a hotspot in this field. Catenary electromagnetics as an emerging metasurface design principle provides new ideas and methods for designing efficient metasurfaces. Here, we proposed a metasurface absorber based on twisted catenary structure that can achieve efficient spin-selective absorption. The simulated results indicate that the perfect absorption can be achieved for left-handed circularly polarized incidence at the working wavelength, while the absorption for right-handed circularly polarized incidence is below 22%. The corresponding circular dichroism is larger than 78%. Besides, the physical mechanism for the chiral absorption is analyzed and a promising application for information encryption is also discussed. This work may find potential applications in chiral imaging and chiral sensing.

Overview: Spectral imaging detection technology has been widely used in many fields, such as remote sensing, medical diagnosis, food safety testing, environmental monitoring, and other fields due to its advantages of accurate and non-contact detection. However, conventional spectral imaging systems usually suffer from the large volume, long sampling time, and low energy efficiency. Metasurface is an artificial two-dimensional material that can flexibly control the amplitude, phase and spectrum of electromagnetic waves. Metasurfaces have been used in spectral detection, holography, metalens, and other fields due to its compact structure and the capacity to flexibly control the electromagnetic waves. Benefiting from the advantages of small size, compact structure, and easy integration, miniature spectral detection technologies based on metasurfaces have been widely studied in recent years. The miniature spectral detection systems usually utilize the broadband spectral properties of metasurfaces and compressive sensing algorithms to achieve computational spectral imaging detection with lightweight. However, the existing designs of the metasurfaces-based miniature spectral detection system usually lack the quantitative analysis of the relationship between the average correlation values of the metasurfaces transmission spectra and the reconstruction quality. The random selection method used in the existing design process cannot guarantee the optimal reconstruction quality. Different from the traditional methodology of using the maximum linear independence criterion to select the broadband filters, this paper quantitatively analyzes the relationship between the average correlation value of the metasurfaces transmission spectra and reconstruction quality, and proposes a methodology for miniature spectral detection based on metasurfaces, which provides a route for the subsequent design and optimization of the metasurfaces. In order to verify the advantages of the proposed methodology, ten broadband spectra and image spectra were selected from many spectra. Compared with the random selection design methodology, the proposed methodology can improve the reconstruction fidelity of broadband spectral and image signals. The fidelity of the broadband spectral reconstruction can be increased by 13.17%, and the reconstruction fidelity of the image spectral signals has also been improved to a certain extent. In addition, this paper also verifies the spectral properties of the metasurfaces-based miniature spectral detection technology, showing that the system has good reconstruction effect for broadband, narrowband and image spectral signals, and has the advantages of compact structure and small volume.Benefiting from the advantages of small size, compact structure, and easy integration, miniature spectral detection technologies based on metasurfaces have been widely studied in recent years. However, the existing designs of the metasurfaces-based miniature spectral detection system usually lack the quantitative analysis of the relationship between the average correlation values of the metasurfaces transmission spectra and the reconstruction quality. The random selection method used in the existing design process cannot guarantee the optimal reconstruction quality. This paper quantitatively analyzes the relationship between the average correlation value of the metasurfaces transmission spectra and reconstruction quality, and proposes a design methodology for miniature spectral detection based on metasurfaces. In addition, this paper also verifies the spectral properties of the metasurfaces-based miniature spectral detection technology. Compared with the random selection design methodology, the proposed methodology can improve the reconstruction fidelity of broadband spectral and image signals.

Overview: Surface plasmon polaritons (SPPs) directional excitation is the basis for the development of on-chip integrated photonic systems, such as the super-resolution imaging, the nano lithography, and the high sensitivity biosensors. It is difficult for traditional directional structures, such as prisms, nano slits and grooves to satisfy the accurate phase-matching condition required for SPPs excitation, resulting in an unsatisfactory coupling efficiency, a low extinction ratio, and high loss and noise. In recent years, the directional excitation of surface plasmon polaritons based on the catenary metasurface began to be valued because of the continuous and linear geometric phase control ability. However, the research of SPPs directional excitation with linearly polarized light is less than that of circularly polarized light. In this paper, all excitation is explained according to the multi-level scattering theory and the Huygens-Fresnel principle. The simulation results show that at the resonance wavelength (836 nm), the SPPs directional excitation is effectively achieved due to the stronger electric dipole excited by SPPs resonances. At the same time, in the effective bandwidth range (820 nm~870 nm) of unit catenary nanoparticle, the electric dipole scattering intensity and spectral extinction ratio curve both show the trend of increasing first and then decreasing. Therefore, there is a positive correlation between the electric dipole scattering intensity and spectral extinction ratio curve. The above analysis shows that the dipole intensity is the main factor affecting the directional extinction ratio. The designed directional excitation of surface plasmon polaritons with linearly polarized light based on the catenary nanoparticle metasurface is effective. The peak extinction ratio is up to 27 dB (corresponding to the incident wavelength of 820 nm), and the bandwidth above 10 dB is about 47 nm (798 nm~845 nm). Therefore, these results are helpful for the research and development of the catenary multifunctional devices which has great potential in the design of SPP directional excitation devices. Besides, it is also a planar integrated device, which can provide new ideas for chip-level photonic device or system design. Moreover, the method in this paper is also suitable for circularly polarized light, therefore it can be referenced in the design of other multi-functional integrated photonic devices such as multi-directional beam splitters and polarization detectors.Surface plasmon polaritons (SPPs) directional excitation is the basis for the development of on-chip integrated photonic systems. And SPPs directional excitation based on the catenary metasurface is a hot and frontier field in recent years, however, the SPPs directional excitation with linearly polarized light is less than that of circularly polarized light. In this paper, we design a catenary nanoparticle metasurface to realize the SPPs directional excitation with linearly polarized light. The spectral extinction ratio curve and electric field distribution under the incident of x-polarized light are calculated with the finite difference time domain. The physical mechanism of SPPs directional excitation is explained according to the multi-level scattering theory and the Huygens-Fresnel principle. The simulation results show that the SPPs directional excitation with linearly polarized light based on the catenary nanoparticle metasurface is effective, and the peak extinction ratio is up to 27 dB (corresponding to the incident wavelength of 820 nm), and the bandwidth above 10 dB is about 47 nm (798 nm~845 nm). Therefore, these results are helpful for the research and development of the catenary multifunctional devices.

With the rapid development of lithium-niobate-on-insulator (LNOI) thin film technology and related surface micro-nano manufacturing technology in recent years, a series of high-quality and high-performance photonic functional devices on lithium niobate chip have been realized, such as compact modulators with ultra-high performance, broadband frequency combs, as well as high-efficiency optical frequency converters and single-photon sources. Great progress has been made in nonlinear optical frequency conversion, electro-optic modulation and optical passivity. In this paper, we briefly introduce several micro-nano processing technologies that have the potential to produce high-quality lithium niobate metasurfaces, and summarize the recent research progress in optical frequency conversion, electro-optic modulation, optical passivity and other aspects of lithium niobate metasurfaces, and prospected the potential research directions in the field of micro-nano optics.As derivatives of 3D metamaterials, artificial metasurface structures with sub-wavelength thicknesses can flexibly manipulate light-matter interactions in a compact manner, which is beneficial for the fabrication of multi-functional and ultracompact photonic devices. Therefore, metasurface structures are of great significance for micro-nano photonics and integrated photonics. The ferroelectric crystal lithium niobate is regarded as one of the most promising multifunctional integrated photonic platforms due to its wide transparent window spanning from the visible to the mid-infrared band as well as large nonlinear optical and electro-optic coefficients. In recent years, research on integrated photonics devices based on lithium-niobate-on-insulator (LNOI) thin films has also been developed rapidly. In this paper, several micro-nano processing technologies that have the potential to prepare high-quality lithium niobate metasurfaces are summarized. At the same time, the research progress of lithium niobate metasurface structures in recent years is introduced, and its future research directions are prospected.

In this paper, a tunable extrinsic chiral metasurface with giant circular dichroism in the optical frequency band is proposed. The unit cell of the metasurface consists of two symmetrical square silver split rings and a GST film sandwiched between them. Compared with the works reported in the existing literature, the CD of the metasurface is larger and the tuning range is wider. In the frequency range of 50 THz~300 THz, the CD of this extrinsic chiral metasurface is up to 0.85 when the GST is amorphous. Due to the large loss of the crystalline GST, the extrinsic chiral response is weakened and the maximum CD is 0.52 in the crystalline state. The GST switches between two phase states (amorphous-crystalline) and enables the frequency tuning range to reach about 70 THz. Further studies have shown that the CD can be tuned by changing the incident angle and the geometric parameters of the GST layer. When θ = 0° and φ= 0°, the CD is zero in all frequency bands, which means when the wave is incident normally, there is no intrinsic chiral characteristics when the wave is incident normally. The electric field distributions at the resonance point in different phase states are also investigated. This work provides a way to realize devices such as efficient polarization modulation devices, circular polarizers, and polarization filters in the optical frequency band.We propose a metasurface with tunable circular dichroism extrinsic chiral based on the phase change material Ge2Sb2Te5 (GST). The metasurface is composed of two symmetrical square silver split ring resonators and a GST intermediate layer arranged periodically. Combined with the oblique incident light, this metasurface is capable of achieving electromagnetic properties similar to that of the chiral structure. Numerical simulation results show that in the frequency range of 50 THz~300 THz, the maximum circular dichroism (CD) of the metasurface is 0.85 in the amorphous state of GST and 0.52 in the crystalline state of GST. When the GST switches between two states (amorphous-crystalline), a tunable frequency of about 70 THz is achieved. Compared with the reported works, the CD of this metasurface is larger and the tuning range is wider. By studying the electric field distribution, the origin of the circular dichroism is explained; and the effects of incidence angles and structural parameters on the CD of this metasurface are investigated. This research will have potential applications in efficient polarization modulation devices, circular polarizers, and polarization filters in the optical frequency band.

Spectral computed tomography (CT) based on photon-counting detectors, has great potential in material decomposition, tissue characterization, lesion detection, and other applications. During the reconstruction, the increase of the number of channels will reduce the photon number in a single channel, resulting in the decline of the quality of the reconstructed image, which is difficult to meet the actual needs. To improve the quality of image reconstruction, joint multi-channel total generalized variational based on the unclear norm for spectral CT reconstruction was proposed in this paper. The algorithm will extend total generalized variation to the vector, and the sparsity of singular values is used to promote the linear dependence of the image gradient. The structural information of the multi-channel image is shared during the image reconstruction process while unique differences are preserved. Experimental results show that the proposed algorithm can effectively recover image details and marginal information while suppressing noise.

A tunable nanosecond pulse fiber laser is demonstrated in the paper. The laser adopts the passive mode locking mechanism of the nonlinear amplifying loop mirror and a manually adjustable filter and fiber grating are added to achieve single-wavelength spectral output. The passive mode locked erbium-doped fiber laser with 430 m cavity length generates the nanosecond rectangle pulse at 465 kHz repetition rate. The tunable passive mode locked fiber laser incorporates a broad bandwidth mode locking device and a tunable filter in the cavity. The broad bandwidth mode locker is the key device for the tunable pulse output, which is based on a reflective nonlinear amplifying loop mirror. The result shows that the pulse duration and the single-pulse energy are 10.58 ns and 70.28 nJ respectively when the laser works at 1560 nm and has 400 mW pump power. The tunable range is from 1523.4 nm to 1575 nm.

As an essential resolution enhancement technique, source optimization can improve the quality of advanced lithography. In the field of advanced lithography, the convergence efficiency and optimization ability of the source optimization are very important. Particle swarm optimization (PSO) is a global optimization algorithm. The adaptive control strategy can improve the global search ability of particles, and the nonlinear control strategy can expand the search range of particles. In this paper, a PSO algorithm based on adaptive nonlinear control strategy (ANCS) is proposed to solve the problem of source optimization by transforming it into a multivariable evaluation function. The image optimization simulation is carried out with a brief periodic grating image and an irregular image, and the source shape is optimized by the global iteration property of the proposed method. By using the pattern errors (PEs) as a multivariate merit function, the results of 300 iterations are evaluated, and the PEs of the two kinds of simulation patterns are reduced by 52.2% and 35%, respectively. Compared with the traditional PSO algorithm and genetic algorithm, the proposed method not only improves the imaging quality, but also has higher convergence efficiency.

An electro-optic intensity chaotic communication system is designed by combining two electro-optic delay feedback loops with parallel structures. By injecting chaos into chaos, a more complex chaotic waveform is generated to enhance the chaotic complexity and the communication system confidentiality. In this design, MATLAB and OptiSystem software are used to simulate the system, which solves the problem that OptiSystem software can't simulate the optical feedback loop. The mature laser and binary sequence generation modules in OptiSystem software provide energy and input signals for the system. The electro-optic delay feedback loops are realized by the MATLAB program, and the signal transmission in optical fibers is completed in the OptiSystem software. The article introduces how to use MATLAB and OptiSystem software to realize the co-simulation of chaotic systems. Numerical simulations show that the proposed method is feasible to simulate the optical feedback loop, and the simulation results are in good agreement with the theoretical values, which prove that the chaotic signal is generated.

Although the current laser active detection system used to find "cat-eye" target has large transmitting power and long detection distance, it generally has the disadvantages of high quality and poor flexibility. In order to enhance flexibility, reduce operational response time and ensure that the target can be destroyed as soon as it is found, a low-power active "cat-eye" detection system can be integrated in the intelligent sight to complement the existing system. Because the intelligent sight is integrated with a small laser rangefinder and CMOS image sensor, according to its hardware characteristics, this paper designs an anti-interference "cat-eye" target detection method, which uses low-power laser to emit pseudo-random coded laser pulse sequence, CMOS sensor to synchronously collect data, and extract target information through correlation operation, and carries out theoretical analysis and experimental verification. The experimental results show that the method has strong stability and anti-interference ability, and can make the intelligent sight find "cat-eye" target in complex background.

In this paper, based on TDLAS technology, an all-fiber NH3 concentration detection system was built by using the designed microfiber gas absorption cell. The core part of the NH3 detection system was sensed by a 1.51-m microfiber. The test results of the system indicate that there is a good linear relationship between the second harmonic amplitude and the corresponding concentration for NH3 in the concentration range of 20000 ppm-100000 ppm (correlation coefficient of fitting formula R=0.9962). To improve the detection performance of NH3 concentration, the gold-nanosphere (GNS) coated microfiber is used to enhance the evanescent field effect. According to the experimental results, the sensitivity of the microfiber coated GNSs NH3 concentration detection system has been greatly improved and the lower detection limit of NH3 concentration can reach 260 ppm. Repeated monitoring of different concentrations of NH3 shows that the detection system is stable with a maximum relative error of 5.38%, which makes it suitable for long-term stable NH3 monitoring and has wide application prospects.

A deep UV tunable narrow-passband light source module for the nitrate measurement system is proposed and demonstrated in this paper. The module consists of a deuterium lamp, an angle rotation stage, a motorized filter wheel, the deep UV optical filters with different central wavelengths, as well as the collimating lens. Seven UV filters with different central wavelengths of 220 nm, 230 nm, 240 nm, 250 nm, 260 nm, 270 nm and 280 nm are placed on the filter wheel. Based on the principle of multiple-beam interference, the central wavelength of the transmission light and the rotation angle are regressed and calibrated to obtain the relationship model. The experiment results demonstrate that with the rotation angle from 0 to 30°, each filter can realize wavelength tuning range of 10 nm. In addition, the designed deep UV tunable light source module can obtain monochromatic light with the wavelength ranging from 212 nm to 280 nm, which meets the measurement requirements of the nitrate in seawater.measurement

The strong localized plasmon resonance of metasurfaces makes the resonance frequency extremely sensitive to the dielectric environment, which can be applied to label-free environment detection. In this paper, a bow-tie terahertz metasurface with an optimized ratio of the quality factor to effective mode volume(Q/Veff) is designed. The unit cell of the proposed structure is composed of a mirror-symmetrical metallic bow tie in the middle and continuous metallic strips on both sides. The width of each metal strip and the length of the bow-tie gap are optimized for the parameter Q/Veff. When the metal strip width is 25 μm and the gap length is 2 μm, the effective mode volume is 3.6 μm3 and Q/Veff is 2.2 μm-3 at 0.7 THz. In the experiment, different concentration of the lead ion solution was dropped on the proposed metasurface. The transmission spectrum was measured by a terahertz time-domain spectroscopy system. The results showed that there is a linear relationship between resonance frequency shift and lead ion solution concentration from 0.1 ng/mL to 20 ng/mL. The detection limit is 0.1 ng/mL. The terahertz metasurface sensor has the advantages of the miniaturized size, easy sample preparation, fast measurement capability and real-time detection, which will be widely used in environmental protection and food safety.(2020190189)

Due to the small scale and weak energy of the infrared dim small target, the background must be suppressed to enhance the target in order to ensure the performance of detection and tracking of the target in the later stage. In order to improve the ability of gradient reciprocal filter to suppress the clutter texture and reduce the interference of the residual texture to the target in the difference image, an adaptive gradient reciprocal filtering algorithm(AGRF) is proposed in this paper. In the AGRF, the adaptive judgment threshold and the adaptive relevancy coefficient function of inter-pixel correlation in the local region are determined by analyzing the distribution characteristics and statistical numeral characteristic of the background region, clutter texture, and target. Then the element value of the adaptive gradient reciprocal filter is determined by combining the relevancy coefficient function and the gradient reciprocal function. Experimental results indicate that the sensitivity of the AGRF algorithm to the clutter texture is significantly lower than that of the traditional gradient reciprocal filtering algorithm under the premise of the same target enhancement performance. Compared with the other nine algorithms, the AGRF algorithm has better signal- to-noise ratio gain (SNRG) and background suppress factor (BSF).助项目(2019AC20147)

Multi-line LiDAR has a wide range of application demands, but the current detection and processing circuit of LiDAR is mostly composed of discrete components, and the detector is separated from the processing circuit, which brings high cost, poor reliability, and other problems. To solve the above problems, an integrated 16-element LiDAR analog front-end micromodule based on system-in-package technology is proposed, which has important practical significance for the research of multi-component LiDAR micromodule. This module integrates a 16-element avalanche photon diode array detector, a self-developed multi-channel LiDAR analog front-end readout circuit chip, a temperature sensor, and a thermoelectric cooler, etc., which can realize the integration of detection, processing and temperature control. The test results show that the thermostatic stability of the micromodule is 0.07 ℃, the bandwidth of the micromodule is up to 190 MHz, the noise level of the integrated micromodule is reduced by more than 32% compared with that of the non-integrated micromodule, and high speed detection of 5 ns laser narrow pulse is realized.

The master, which contains the desired optical surface, is epoxied to the substrate. When the pieces are separated, the epoxy resin layer is transferred to the substrate producing a replicated mirror. Epoxy replication is an efficient and low-cost way to fabricate optical mirrors. The surface figure accuracy will decrease with the increase in mirror size due to the characteristic of the epoxy resin, and there is no effective way to correct the surface figure aberrations after replication. Finite element analysis was used to simulate the replication process and optimize the thickness of the master for better surface figure accuracy. A multilayer film compatible with Magneto-Rheological Finishing was also developed. A parabolic replicated mirror with a diameter of Φ180 mm and a plane replicated mirror with a diameter of Φ500 mm were fabricated within 5 and 10 days, respectively. The precision shape (RMS<20 nm) and low surface roughness (Rq=0.6 nm) were both achieved.

Defocus dithering technology effectively avoids the non-linear effects introduced by commercial projectors and can achieve high-speed dynamic measurement. However, the binary dithering stripes after defocusing and the standard sinusoidal stripes are not completely close, and there will be a certain deviation in the measurement process. Aiming at the measurement error, this paper proposes an optimization method based on an improved binary ant colony algorithm, which optimizes and improves the binary defocus dithering technology from the phase domain to better improve the measurement accuracy. Through the optimization of small-size binary blocks, the entire optimization process is replaced, and the optimization efficiency is improved. Simulation and experimental results prove that this method can achieve effective three-dimensional measurement under different defocus conditions.

Aiming at the problems of low signal-to-background ratio and low wavefront detection accuracy of the adaptive optical system Hartmann sensor under strong background conditions during the day, based on the difference in polarization characteristics between man-made targets and strong backgrounds, a polarized Hartmann wavefront detection technology is proposed. The traditional Hartmann wavefront detection is converted from the intensity dimension to the polarization dimension, thereby effectively improving the signal-to-background ratio and wavefront detection accuracy. The basic methods and principles of polarized Hartmann wavefront detection are described, and the correctness and accuracy of the method are verified through numerical simulations and experiments. Theoretical and numerical simulation results show that the polarized Hartmann wavefront detection technology can effectively improve the signal-to-background ratio and wavefront detection accuracy under strong background conditions, and significantly enhance the ability of the adaptive optical system to work under strong background conditions.

To prevent people from using the processed turquoise and counterfeit turquoise in medicine, this paper focuses on identifying the raw materials of medicinal turquoise. A turquoise identification system was developed using hyper-spectral imaging technology. The sample standard spectral line was obtained while the applicability was analyzed by the present sample, based on the high-resolution spectral data of ore samples from 6 representative producing areas of natural turquoise in China. A new method was summarized by the differences in correlation coefficients in the range of 400 nm?1000 nm and 400 nm?600 nm of the fake turquoise on the market and an experimental prototype system to identify the true or false of turquoise was developed. Further research will provide technical support to select raw materials in mineral medicine, which will greatly promote the modernization of Tibetan medicine.

Artificial knee joint plays an important role in improving the joint condition of patients. The surface deviation of artificial knee joint will directly affect the treatment effect of patients. Therefore, it is necessary to evaluate the surface of artificial knee joint with high precision before it is put into use. The slope of the artificial knee joint is complex and varies greatly, and there are great differences in the knee joint surface among different patients. The complexity and unknowability of artificial knee joint make it difficult to measure its surface with high precision. In this paper, a normal vector tracking measurement method based on contact inductance linear displacement sensor (LVDT) is proposed, and a rotary scanning measurement system is built. This method performs curve fitting on the measured points, predicts the changing trend of the measured surface, and adaptively adjusts the sampling position and posture of the LVDT to make it measure approximately along the normal direction of the sampling point, so as to realize the adaptive rotation measurement of the complex and unknown surface with large slope. Through the measurement experiment of the standard ball, the measurement error of the calibration system is about 48.21 μm. In addition, the measurement experiment of the artificial knee joint model is carried out to verify the feasibility of the method.

We describe a modified six-step method to simultaneously measure the inhomogeneity of sample plate and the planarity of the four surfaces in an absolute manner, along with a high-efficiency iterative algorithm for data reduction. Combined with the iterative algorithm, the errors of inhomogeneity and flatness can be estimated with pixel-level spatial resolution in a fast and effective manner. The simulation and experiments prove the validity of the method and the measurement capability of reaching sub-nanometer accuracy. The method presented in this paper is cross-compared with traditional absolute testing method and the method of inhomogeneity. The difference between absolute plane measurements is less than 1.7 nm RMS, and the difference of inhomogeneity measurement accuracy is less than 2.3 nm RMS. The experimental results show that these two methods are highly consistent and have good repeatability, which verifies the accuracy of the methods proposed in this paper. Uncertainty analysis indicates that the proposed method improves the measurement uncertainty, compared with the classical transmission method.

A light field camera can simultaneously sample a scene from multiple viewpoints with a single exposure, which has unique advantages in portability and depth accuracy over other depth sensors. Noise is a challenging issue for light field depth estimation. Most of the traditional depth estimation methods for noisy scenes are only suitable for non-occluded scenes, and cannot handle the noisy scenes with occluded regions. To solve this problem, we present a light field depth estimation method based on inline occlusion handling. The proposed method integrates the occlusion handling into the anti-noise cost volume, which can improve the anti-occlusion capability while maintaining the anti-noise performance. After the cost volume is constructed, we propose a multi-template filtering algorithm to smooth the data cost while preserving the edge structure. Experimental results show that the proposed method has better performance over other state-of-the-art depth estimation methods in high noise scenes, and can better handle the occlusion problem of depth estimation in noisy scenes.

The tibial shaft fracture model was customized by reverse engineering and 3D printing technology, and the biomechanics of the Orthofix Unilateral External Fixator for tibial shaft fracture was studied. Through the design of an orthogonal experiment scheme, the distribution of the Schanz’s nails on the clamp, the distance from the lateral the Schanz’s nails to the fracture end, and the distance from the tibia to the external fixture were measured by the XTDIC-CONST 3D Full-Field Strain Measurement and Analysis System. The experimental results show that when the number of the Schanz’s nails decreased, the bending deformation of the Schanz’s nails will increase from pressure load, which increases the possibility of plastic deformation and fatigue fracture of the external fixator. According to the mechanical analysis results of the nine schemes, the distance from the external fixture to the tibia has the most significant effect on the deformation of the Schanz nail. When installing six Schanz pins in the clip, the distance from the lateral Schanz’s nail to the fracture end is 120 mm, and the distance from the external fixture to the tibia is 30 mm. The comprehensive performance of the scheme is the best.

During the preparation and use of the optical thin film, the absorption center will be generated due to defects and pollution. When the optical thin film is irradiated by a laser, the absorption center absorbs light energy and generates thermal signals, according to which the optical absorption loss of an optical film can be measured. The method proposed in this paper for measuring the optical absorption loss of a thin film based on a thermal imager. The addition of a reference sample in the test can reduce the impact of the changes of environmental temperature and the thermal imager noise on the temperature test results. Taking a certain area of the temperature field recorded by the thermal imager during the entire laser irradiation process can reduce the errors of the finite element simulation calculation caused by the laser pointing fluctuations and the unsatisfactory spot distribution. Using this method, the absorption loss of a small-size 45° high-reflection film was tested to be 7.60 ppm, and the spatial distribution of the absorption loss of the same batch of large-size optical film samples were tested. The absorption of the optical film measured by this method is consistent with the result of the laser calorimetry test. This method does not require long-term constant temperature and strict environmental temperature control, and the tested sample size is not limited.

The coal port will produce dust in the process of unloading coal by the chute of the ship loader. In order to solve the problem of dust detection, this paper proposes a method of coal dust detection based on deep learning (YOLOv4-tiny). The improved YOLOv4-tiny network is used to train and test the dust data set of chute discharge. Because the detection algorithm cannot get the dust concentration, this paper divides the dust into four categories for detection, and finally counts the area of detection frames of the four categories of dust. After that, the dust concentration is approximately judged through the weighted sum calculation of these data. The experimental results show that the detection accuracy (AP) of four types of dust is 93.98%, 93.57%, 80.03% and 57.43%, the average detection accuracy (mAP) is 81.27% (which is close to 83.38% of YOLOv4), and the detection speed (FPS) is 25.1 (which is higher than 13.4 of YOLOv4). The algorithm can balance the speed and accuracy of dust detection, and can be used for real-time dust detection to improve the efficiency of suppressing coal dust generated by chute discharge.

Blind image restoration aims to accurately estimate the blur kernel and the wanted clear image with no-reference. Existing researches show that the use of the Total Variation to model the high-order image gradient prior constraints can effectively suppress the blocking artifact generated in the restored image. On the basis of experimental observation and research, this paper proposes to use the sparse prior constraint model to regularize the blind restoration process to obtain a better image restoration performance. Our method makes use of the sparsity of the high-order gradient of the image and combines it with the low-order gradient to construct the mixed gradient regularization term. At the same time, an adaptive factor based on image entropy is introduced to adjust the ratio of the two types of gradient priors in the iterative optimization process so as to obtain better convergence. Simulated and experimental results prove that compared with the existing state-of-the-art methods of blind image restoration, the proposed method has superior image restoration performance.

In order to realize the miniaturization and integration design of space optical communication system, an integrated tracking system based on the array detector and the fast deflection mirror is established. By analyzing the principle of spot position detection of array detector, a centering algorithm is proposed. Firstly, the coarse centering strategy is designed by setting the threshold value. Then, the fine centering is completed by using the database query method. Thirdly, the infinite integral method is used to make the spot return to the origin. Finally, the correctness and feasibility of the algorithm are verified by building an experimental platform. The experimental results show that the tracking field of view can reach 70.3 mrad, which is about 3 times larger than that of the original algorithm, and the maximum tracking error is better than 1.8 μrad, which lays a foundation for further engineering application of the space optical communication system.

In order to respond to the scintillator screening requirements of large scientific projects and the new medical imaging equipment such as the development of large-scale collider experimental detectors, space load calorimeters and TOF-PET, our laboratory conducts research on the scintillation performance (emission spectrum, light output, energy resolution, decay time, afterglow, coincidence time resolution, etc.) of scintillators. A complete set of inorganic scintillator performance test programs is designed for the optimal performance of different scintillator samples. In the test of emission spectrum, different excitation sources were selected for comparison test. The energy resolution and the test conditions of the scintillation performance such as time resolution were optimized, which were successfully applied to the performance research of popular scintillators including cerium-doped yttrium lutetium silicate (LYSO:Ce) and gadolinium aluminum gallium garnet (GAGG:Ce), and good test results were obtained. The energy resolution of LYSO:Ce and ceramic GAGG:Ce scintillators are 7.9% and 5.4%, respectively, and the coincidence time resolution of the LYSO:Ce scintillator can reach 94.3 ps.

To solve the problem of motion blur in abnormal behavior detection, a fast motion blur removal algorithm, based on DeblurGAN, is proposed. Three 3×3 convolutions are used to replace the 7×7 convolution in the original generator. The transposed convolution is discarded. Firstly, bilinear interpolation is used to expand the size of the feature map which needs upsampling. The residual unit is replaced by a residual density block (RRDB) in the original algorithm. The RRDB is then scaled to 0~1 to avoid unstable training. The L1 loss of gradient images is added to the loss function of the original generator. As the DeblurGAN reconstructed image edge is often not clear enough, the edge information of the image is added to make the reconstructed image edge more obvious. The effectiveness of this method is verified by experiments and is compared with other similar algorithms like DeblurGAN. The PSNR of the optimized model is improved by 0.94. The structure similarity and speed are equivalent. The chessboard lattice problem in the reconstructed image is solved. The edge of detail is more prominent. The performance of the proposed model is better than that of other related algorithms.

The liquid crystal optical phased array (LCOPA) is the core device of next-generation beam control technology. Improving its laser-induced damage threshold is one of the current research hot spots. Aiming at the scene of higher power laser incidence, the degradation degree of LCOPA phase modulation performance should be evaluated. Based on the traditional quarter-wave plate method, this paper realizes fast and direct measurement of the phase modulation of the liquid crystal to the incident laser. The verification test found that when the core temperature is 33 ℃, the corresponding maximum phase aberration is 3.6 rad. At the same time, based on the measured phase modulation results, this paper studies the deterioration process of the beam quality of the outgoing light. Analysis results show that the deterioration of beam quality is less than 20% when the core temperature of the liquid crystal phase shifter changes less than 10 ℃.

With the development of full-screen mobile phones, the need for under-screen imaging of mobile phones has emerged. However, the diffraction caused by the wiring and other opaque parts will affect the image quality of the under-screen image. In this article, under-screen image is restored from the perspective of image restoration. The point spread function (PSF) of the mobile phone imaging system is obtained through actual measurement, and the image is deconvolved using the measured PSF. In this article, traditional deconvolution method has been improved, in which the color space of the image is converted and different channels are processed separately. Compared with the traditional deconvolution method, the results of the sub-channel deconvolution method have improved structural similarity (SSIM), peak signal-to-noise ratio (PSNR) and other indicators, and the required running time is shorter. After sub-channel deconvolution, the non-local averaging algorithm is used for denoising, which further improves the quality of the under-screen image.

To solve the problems of weak signal strength and low signal-to-noise ratio in traditional Raman spectroscopy, a new confocal Raman system is proposed in this paper. The absolute conjugation of the confocal point is realized by external photonic crystal fiber. The technical problems in the coupling process of photonic crystal fiber are summarized, and the actual samples are tested. Compared with conventional confocal Raman fibers such as Thorlabs and OZ and Witec 532 nm-alpha300R Raman system, the signal-to-noise ratio in this paper is 73.8382 at the same laser intensity and integration time, which is significantly higher than that of Thorlabs and OZ (37.1557 and 40.0342, respectively). Compared with the signal-to-noise ratio of 65.5312 for Witec 532 nm-alpha300R, it also increased by 12.68%. High-quality Raman signal makes the absolute conjugated confocal Raman system have broad market prospects and ultra-high market competitiveness.

In order to improve the nonuniformity caused by the process bias of sCMOS readout circuit, an adaptive multipoint nonuniformity correction method is presented. The algorithm first determines the location of the optimal segment point and the optimal number of segments by searching for the minimum norm point and threshold comparison, then corrects two points in each interval segment according to the segment information. This adaptive method can effectively improve the correction performance of traditional multipoint methods, which is caused by improper selection of segment parameters. At the same time, in order to achieve real-time non-uniformity correction, a matching embedded data series correction scheme is proposed based on the algorithm characteristics of adaptive multipoint method, which can achieve non-uniformity correction without affecting the existing camera acquisition structure and acquisition rate.

Apodization has found many important applications in imaging and optical communication. Traditional apodization methods are based on the phase or amplitude modulation, suffering from either narrow working bandwidth, or reduced spatial resolution. Here, a broadband achromatic metasurface filter is proposed to realize apodization imaging without sacrificing the spatial resolution. With this filter, a nearly dispersionless phase modulation in the entire visible waveband can be achieved. The simulated results indicate that the focusing efficiency of the metasurface filter is twice larger than that of the phase filter and the imaging contrast can be improved by three times with the metasurface filter compared to the Gaussian filter. The sidelobes in the point spread function can also be efficiently suppressed to the scale of 10-5 in the whole visible spectrum ranging from 400 nm to 700 nm with our design. Additionally, the resolution of diffraction limit or even sub-diffraction can be achieved with this method.

Addressing on the issues like varying object scale, complicated illumination conditions, and lack of reliable distance information in driverless applications, this paper proposes a multi-modal fusion method for object detection by using convolutional neural networks. The depth map is generated by mapping LiDAR point cloud onto the image plane and taken as input data together with the RGB image. The input data is also processed by the sliding window to reduce information loss. Two feature extracting networks are used to extract features of the image and the depth map respectively. The generated feature maps are fused through a connection layer. The objects are detected by processing the fused feature map through position regression and object classification. Non-maximal suppression is used to optimize the detection results. The experimental results on the KITTI dataset show that the proposed method is robust in various illumination conditions and especially effective on detecting small objects. Compared with other methods, the proposed method exhibits integrated advantages in terms of detection accuracy and speed.

In the task of person re-identification, there are problems such as difficulty in labeling datasets, small sample size, and detail feature missing after feature extraction. The joint discriminative and generative learning for person re-identification of the deep dual attention is proposed against the above issues. Firstly, the author constructs a joint learning framework and embeds the discriminative module into the generative module to realize the end-to-end training of image generative and discriminative. Then, the generated pictures are sent to the discriminative module to optimize the generative module and the discriminative module simultaneously. Secondly, according to the connection between the channels of the attention modules and the connection between the attention modules in spaces, it merges all the channel features and spatial features and constructs a deep dual attention module. By embedding the models in the teacher model, the model can better extract the fine-grained features of the objects and improve the recognition ability. The experimental results show that the algorithm has better robustness and discriminative capability on the Market-1501 and the DukeMTMC-ReID datasets.

With the development and application of blue semiconductor lasers, it has become a research hotspot to obtain high brightness blue light source by beam combining technology. In order to obtain high brightness blue light output, 48 single tube semiconductor lasers with wavelength of 450 nm and output power of 3.5 W are focused and coupled into 105 μm/0.22 NA fiber by fast slow axis collimation and spatial beam combination. The blue light with power of 144.7 W and brightness of 11 MW/(cm2?str) is obtained. The coupling efficiency is 93.78%, and the optical to optical conversion efficiency of the whole system is 86.13%.

An all-silicon PIN photodetector based on black silicon microstructure is reported. The device combines the characteristics of broad spectrum and high absorption of black silicon structure and the characteristics of high quantum efficiency and high response speed of PIN photodetectors. By adding a black silicon microstructure layer based on the traditional silicon PIN photodetector structure, the response characteristics of the detector in the near-infrared band are improved without affecting the response speed. A method is proposed to solve the contradiction between quantum efficiency and response speed in the vertical structure of the PIN photodetector. The test results show that the quantum efficiency of the device can reach 80%, and the peak wavelength is 940 nm. The light responsivity reaches 0.55 A/W, and the dark current is about 700 pA. The response time is 200 ns.

We demonstrate a new mode-locking method: multimode interference mode-locking. This method is simple and convenient in construction. It is only necessary to fuse two short pieces of graded-index multimode fiber in a single-mode fiber laser, which uses the mode interference effect of single-mode multimode single-mode (SMS) structure to achieve saturable absorption mechanism. In order to realize the mode-locking of the SMS structure, it is necessary to precisely control the length of multimode fiber. We propose to coil the SMS structure into the polarization controller. By theoretically deriving the polarization controller to adjust the phase of transmission light in a multimode fiber, the saturable absorption effect can be achieved. Under the 263 mV pump power, a stable 24.83 MHz repetition frequency fundamental frequency mode-locked pulse output was realized, where the pulse interval was 40.12 ns, the signal-to-noise ratio was 50.8 dB, and the center wavelength was 1881.7 nm. The conversion between soliton molecules and traditional soliton can be realized by adjusting the polarization controller and pump power. Under the pump threshold of 410 mW, a stable 25 MHz repetition frequency soliton molecular mode-locked pulse output was realized, where the pulse interval was 40.3 ns, the signal-to-noise ratio was 54.4 dB, and the center wavelength was 1887.60 nm.

A stacked liquid lens based on electrowetting-on-dielectric (EWOD) is designed to analyze the ability of correcting the distorted wavefront caused by curvature, tilt, and piston. The model of the stacked liquid lens is constructed by COMSOL software which is used to simulate the change of liquid interface with different voltage combinations, and the change range of the interface. The correction ability of the stacked liquid lens at a certain point in the wavefront is assessed from wavefront image and point spread function (PSF) is got by ZEMAX software. The results show that different types of distorted wavefront can be compensated via the stacked liquid lens. The peak-to-valley (PV) value decreases from 19.7853? to 0.18?, and the root mean square (RMS) value is down from 5.6638? to 0.0355?. Concurrently, the Strehl ratio (SR) increased from near 0 to 0.962. The related research results have broad prospects in the field of wavefront correction.

Atmospheric polarization has broad application prospects in navigation and other fields. However, due to the limitation of the physical characteristics of the atmospheric polarization information acquisition device, only local and discontinuous polarization information can be obtained at the same time, which has an impact on the practical application. In order to solve this problem, by mining the continuity of atmospheric polarization mode distribution, this paper proposes a network for generating atmospheric polarization mode from local polarization information. In addition, polarization information is often affected by different weather conditions, geographic environment and other factors, and these polarization data are difficult to collect in the real environment. To solve this problem, this paper mines the diversity relationship between the few-shot data under different weather and geographic conditions, by which the generated atmospheric polarization mode is generalized to different conditions. In this paper, experiments are carried out on the simulated data and measured data. Compared with other new methods, the experimental results prove the superiority and robustness of this proposed method.

In order to achieve effective and reliable video transmission, a video joint coding scheme based on dictionary learning and the concatenation of LT code and LDPC code is proposed for underwater single-photon communication system. Sparse coding based on dictionary learning greatly compresses the amount of video data. According to the deletion characteristic of underwater single-photon channel, using the LT-LDPC channel concatenated coding method can overcome the disadvantage of excessive decoding overhead of LT code. Aiming at the problem of decoding failure probability of LT coding, a double feedback mechanism for decoding success is proposed. The experimental results show that when the channel error rate is in the order of 10-2 and the video compression rate is 75.6%, the video frames can be reconstructed with an average peak signal-to-noise ratio (PSNR) of 37.4921 dB.

Aiming at the technical difficulties in the rapid detection and reconstruction of three-dimensional micro-nano devices that are difficult to achieve both high precision and high speed, this paper proposes a structured light detection method based on time-domain phase shift technology. The measured light is modulated by a spatial light modulator, and the time-domain phase shift technology is further employed to realize the detection and reconstruction of three-dimensional micro-nano devices. Compared with the traditional structured light detection method, this technology uses the spatial light modulator to measure the phase shift while the sample is scanned axially, so as to ensure the measurement accuracy and improve the measurement efficiency. By analyzing the measurement data, this method can quickly realize three-dimensional shape detection and reconstruction, and the measurement accuracy can be better than 10 nm.

Polyimide (PI) film is widely used in aerospace, microelectronics, and other fields because of its excellent thermal stability and mechanical strength. However, there are very few reports about its application in the direction of optical imaging. To use PI film for imaging, the requirements for the optical homogeneity of the PI film are extremely demanding. The optical homogeneity of the stretch-resistant PI film proposed in this paper with 100 mm diameter and low thermal expansion coefficient meets the Rayleigh criterion, which has the potential for applications in the imaging field. In addition, the tensile strength of this PI is 285 MPa, which is ~2.6 times that of the PMDA-ODA type PI; the coefficient of thermal expansion is about 3.2 ppm?K-1, which is comparable to that of the Novastrat?905 type PI and is one order of magnitude lower than that of the commercial PI films. These excellent basic properties reserve more space to further improve the space adaptability of the PI film. The solution of the optical homogeneity of the PI film will lay the foundation for its application in thin film diffractive optical elements.

Real-time detection of small objects is always a difficult problem in image processing. Based on the target detection algorithm of deep learning, this paper proposed an end-to-end neural network for mobile phone small target detection in complex driving scenarios. Firstly, an end-to-end small target detection network (OMPDNet) was designed to extract image features by improving the YOLOv4 algorithm. Secondly, based on the K-means algorithm, a K-means-Precise clustering algorithm of more appropriate data samples distribution in the clustering center was designed, which was used to generate prior frames suitable for small target data, so as to improve the efficiency of the network model. Finally, we constructed our own data set with supervision and weak supervision, and added negative samples to the data set for training. In the complex driving scene experiments, the OMPDNet algorithm proposed in this paper can not only effectively complete the detection task of using mobile phone while driving, but also has certain advantages over the current popular algorithms in accuracy and real-time for small target detection.

This paper proposes an efficiency-tunable terahertz focusing lens based on the graphene metasurface. The unit cell is composed of two symmetrical circular graphene hollows and an intermediate dielectric layer, wherein the hollow circular middle is connected by a rectangular graphene sheet. This structure can realize polarization conversion, for example, when an incidence with left-hand circular polarization emitted on the metasurface the polarization of the transmitted light is right-hand circular polarization. According to the principle of geometric phase, by rotating the direction of the rectangular bar, the transmitted wave will carry an additional phase and can cover the range of 2π. An THz focusing lens can be realized by properly arranging the unit structure of the graphene metasurface. The simulation results show that the conversion amplitude of circular polarized light can be adjusted by changing the Fermi level of graphene, and the focusing efficiency of the metalens can also be dynamically adjusted. Therefore, this graphene metasurface-based efficiency-tunable focusing lens can be realized by changing the Fermi level without changing the size of the unit cell, and can be widely used in terahertz applications such as energy harvesting and imaging.

The mode field diameter is an important parameter of single-mode fiber, and the GB.15972.45-2008 recommends using the far-field variable aperture method to measure it. This paper analyzes the distribution of the propagating light field in a single-mode fiber. The mode behavior of the light field is the solution of the Helmholtz equation, which in theory should satisfy the Bessel function. In this regard, a method using Bessel function to fit the optical field distribution of the fiber based on the far-field variable aperture method is proposed, and the mode field diameter is calculated from the fitted mode field distribution curve. Compared with the commonly used far-field variable aperture method, when the measurement data is normal, this method has the same measurement accuracy. When there are errors in the measurement data, this method can still ensure the stability and accuracy of the measurement results.

In the phase measuring profilometry, the phase measuring accuracy could be heavily affected by the nonlinearity effects of the projecting and imaging devices. Therefore, it is very important to reduce the nonlinear errors fast and efficiently. An analytic model of nonlinear errors is introduced. Then we propose a phase compensation method which is based on the accurate mathematical model of the phase error. The proportion of each harmonic component is collected by using a large-step phase-shifting algorithm to measure a reference plane. Then the phase errors of the measured object could be compensated by an iterative algorithm. The experimental results show that the proposed method can realize nonlinear error compensation effectively and improve the precision of phase measurement. Meanwhile, since all the harmonic components are pre-calibrated, there is no extra fringe needed, which can meet the requirements of fast and real-time measurement.

The star/airborne optical remote sensing image has a wide field of view and a complex scene. It is easy to produce a large number of false alarms that are similar to the ship's target due to the impact of the shore construction and broken cloud, causing great interference to the ship's detection. Traditional marine ship detection algorithms are difficult to be effective extracting discriminative features that are conducive to detection, results in low detection rates and high false alarm rates for ships. In view of this, this paper proposes an optical ship target detection method combining hierarchical search and visual residual network from the perspective of low false alarm and low missed detection. Firstly, the land and sea area are segmented based on the texture integral map; secondly, the target candidate area is extracted by combining the multi-scale local structural features; then, the primary false alarm is removed by the layered removal strategy based on multi-dimensional visual features; finally, the visual residuals are built the network finely removes false alarms from suspected candidate areas to obtain the final detection result. Based on the GF2 remote sensing GF2 set, the algorithm proposed in this paper is tested and verified. The comprehensive detection rate of this algorithm is 92.0%, the false alarm rate is 12.58%, the average processing time is 0.5 s, the detection effect is good, the efficiency is high, and the adaptability to various scenes is good. It can achieve accurate and efficient detection and positioning of optical ships in complex environments.

Aiming at the working characteristics of the end-effector of the gantry robot with only three translational degrees of freedom, a method for calibrating the point cloud coordinate system of the 3D vision sensor and the tool coordinate system of the robot actuator is designed, on the basis of the traditional two-step method of hand-eye calibration. In this method, only three calibration target pictures and three sets of point clouds are collected by two orthogonal translation movements of the robot, the rotation matrix and translation vector of the hand-eye relationship can be calibrated by measuring the base coordinates of mark points on the target through the TCP contact of the actuator. The method is simple to operate, and the calibration target is easy to make with low cost. The XINJE gantry robot and 3D vision sensor of structured light was used to build an experimental platform for experiments. The results show that the method has good stability and is suitable for field calibration, with calibration accuracy within ?0.2 mm.

Stereoscopic image zoom optimization is a popular basic research problem in the field of image processing and computer vision in recent years. The zoom visual enhancement technology of 3D images has attracted more and more attention. To this end, this paper proposes a method of stereoscopic zoom vision optimization based on grid deformation from the model of camera zoom shooting, and strives to improve the experience of 3D stereoscopic vision. Firstly, use the digital zoom method to simulate the camera model to properly zoom in on the target area, and then establish the mapping relationship between the reference image and the target image according to the camera zoom distance. Secondly, extract the foreground target object and use the modified just noticeable depth difference (JNDiD) model to guide the adaptive depth adjustment of the target object. Finally, combined with the seven grid-optimized energy terms designed in this paper, the image grid is optimized to improve the visual perception of the target object and ensure a good visual experience for the entire stereoscopic image. Compared with the existing digital zoom method, the proposed method has better effects on the size control of the image target object and the depth adjustment of the target object.

Visual tracking algorithm based on a Siamese network is an important method in the field of visual tracking in recent years, and it has good performance in tracking speed and accuracy. However, most tracking algorithms based on the Siamese network rely on an off-line training model and lack of online update to tracker. In order to solve this problem, we propose an online learning-based visual tracking algorithm for Siamese networks. The algorithm adopts the idea of double template, treats the target in the first frame as a static template, and uses the high confidence update strategy to obtain the dynamic template in the subsequent frame; in online tracking, the fast transform learning model is used to learn the apparent changes of the target from the double template, and the target likelihood probability map of the search area is calculated according to the color histogram characteristics of the current frame, and the background suppression learning is carried out. Finally, the response map obtained by the dual templates is weighted, and the final prediction result is obtained. The experimental results on OTB2015, TempleColor128, and VOT datasets show that the test results of this algorithm are improved compared with the mainstream algorithms in recent years and have better tracking performance in target deformation, similar background interference, fast motion, and other scenarios.

A photonic crystal fiber (PCF) for long distance communication was proposed in this paper. The circular and elliptical air holes distribute in the cladding, and there are two small elliptical air holes around the core in cross section of the PCF. The characteristics of the PCF were analyzed by using the finite element method (FEM) systematically. The results show that the PCF offers an ultrahigh birefringence of 3.51×10-2 and the confinement loss as low as 1.5×10-9 dB/m with the optimal structure at the wavelength of 1550 nm. Compared with the existing photonic crystal fibers with elliptical air holes, the birefringence has a large increase, and the confinement loss reduces by 5 orders of magnitude. Additionally, we also analyzed the relationship between the dispersion of the PCF and the wavelength, and obtained the Brillouin gain spectrum characteristics. In general, the PCF can be used in long distance communication system.

The variation of ambient refractive index and ambient temperature is the main factor affecting the error of optical fiber strain measurement. In this paper, a strain sensor based on the dual-mode fiber (DMF) long period fiber grating (LPFG) is designed. The sensor model structure was designed, and the sensor samples with optimized parameters were produced. The experiment tested the response of the DMF-LPFG sensing structure to the strain, temperature and refractive index in the external environment. Through the Bragg grating (fiber Bragg grating, FBG) written on the single-mode fiber with a UV laser, the cross effect of the ambient temperature is solved. The results of the axial strain experiment show that the axial strain sensitivity of the new structure sensor can reach -5.4 pm/με in the strain range of 0 με~840 με, which is greatly improved compared to the ordinary LPFG. The sensitivity is 58.86 pm/℃ in the temperature range of 25 ℃~80 ℃, showing good linearity. At the same time, the sensor is insensitive to changes in ambient refractive index. The dual-parameter matrix is used to process the strain and temperature sensitivity of the few-mode LPFG and FBG to achieve dual-parameter simultaneous demodulation. The new composite grating structure has good sensing performance and engineering application prospects.

In order to reduce the error and improve the measurement accuracy, a more detailed error model is established for the Hartmann method of focal power measurement in this paper. It focuses on the analysis of several problems that cause the error of refraction problems, including the dispersion error of the light source, the inaccurate of the photodetector’s central positing, the tilt of lens, misalignment between incidence axis and main axis of lens, and the incident light and the lens are not perpendicular. At last, it is concluded that the inaccuracy of the center extraction on the photodetector will cause a large error to the final result. For all these reasons, a method of dual bilinear interpolation combined with a fitting method to find the centroid is proposed, proving its effectiveness and accuracy.

In order to realize the non-destructive and real-time dynamic stress monitoring method of the construction machinery surface in complex and harsh environments, a fiber Bragg grating (FBG) stress sensor packaging method based on magnetron sputtering technology is proposed. Two packaging methods of complete embedding (the capillary copper tube embedded in the entire grating area) and two sides embedding (capillary copper tube nested at both ends of the grating area) are studied. The sensitization effect of the sensor is analyzed from the perspective of theory and finite element, and the results are consistent. The physical sensors are made, and temperature, stress, and comparison experiments are carried out. Simulation and experiment show that the FBG sensor improves the sensitivity by about 7.5% under this model. The temperature experiment shows that the temperature feedback correlation coefficient R2 of the second package structure reaches 0.99948, which shows good linearity in the range of 30 ℃?80 ℃; the stress experiment correlation coefficient R2 also reaches 0.99924, and the sensitivity is 6.14 pm/MPa. The accuracy of demodulation system reaches 0.05 MPa, it can demodulate stress quickly and accurately. Comparative experiments show that the monitoring system composed of grating demodulator has higher accuracy than the monitoring system composed of strain gauges, and maximum deviation value smaller 59.8%. The packaging structure of metallization method of embedded capillary copper tube combined with organic glue fixed is simple, high sensitivity, and precision, can meet the needs of large-scale construction machinery surface non-destructive real-time health monitoring.

Array microstructure optical elements are widely used in various beam homogenization occasions, but conventional processing methods cannot meet the accuracy requirements of large-sagittal convex cylindrical arrays. In this paper, the ultra-precision turning forming method is used to analyze the main factors affecting diamond turning, the sequential search method and the binary search method are designed to find the turning track, and the advantages and disadvantages of the two methods are compared. Furthermore, the binary search method is successfully found by combining the Matlab software turning trajectory and the numerical control program. As proof-of-concept demonstrations, turning experiments are carried on an ultra-precision lathe, and a large-vector high-array microstructure with a surface profile error of 135 nm is obtained. It proves that the force binary search method can accurately obtain the turning trajectory, and this method can be applied to both spherical and aspherical contours, showing important engineering application value.

In view of the limitations of the existing methods when the camera has no common field of view, this paper proposes a method of using two plane calibration plates to calibrate two cameras at the same time. By deriving the coordinate transformation between the two cameras and two calibration plates, the solution of the relative pose relationship between any camera and the reference camera is transformed into a more mature hand-eye calibration equation. The experimental results show that this method can achieve simultaneous calibration of two cameras, and the absolute error is less than 0.089 mm. In the dual vision 3D measurement system, the cumulative error with phase height is less than 0.116 mm, which can provide a reliable initial value for the next step of data fusion.

The traditional respiratory rate measurement technologies have several deficiencies, such as subjective appraised results, complicated signal extraction processes, difficult access to equipment, and inconvenience to move due to the wired connection setting. The respiratory airflow can directly reflect the human breath, and the respiratory frequency is usually 10?12 breaths/min (1 ventilation every 5?6 seconds). The humidity difference between exhalation and inhalation can be directly used to measure respiratory rate. In the present work, a wireless respiratory rate monitoring system based on inorganic halide perovskite humidity sensor was developed. The sensor exhibits an ultrasensitive humidity sensing performance, which overcomes the long response/recovery time (> 10 seconds) of the commercial humidity sensors. The system utilized a Zigbee wireless communication to transmit the measurement signal, which separates the signal detection and processing parts, making it easier for the tester to move. The upper computer software was designed and used for data processing to calculate the breathing rate. The system can accurately monitor the respiratory rate in real-time, recognize and alarm the apnea successfully by comparing with a setting threshold value. The test results show that the system can accurately monitor the breathing rate with a maximum error of 1 time per minute. The system possesses great potential for application in respiratory rate monitoring due to its high accuracy, simple operation, portability, and low cost.

The vehicle identification code (VIN) is of great significance to the annual vehicle inspection. However, due to the lack of character-level annotations, it is impossible to perform the single-character style check on the VIN. To solve this problem, a single-character detection and recognition framework for VIN is designed and a weakly supervised learning algorithm without character-level annotation is proposed for this framework. Firstly, the feature information of each level of VGG16-BN is fused to obtain a fusion feature map with single-character position information and semantic information. Secondly, a network structure for both the character detection branch and the character recognition branch is designed to extract the position and semantic information of a single character in the fusion feature map. Finally, using the text length and single-character category information, the proposed framework is weakly supervised on the vehicle identification code data set without character-level annotations. On the VIN test set, experimental results show that the proposed method realizes the Hmean score of 0.964 and a single-character detection and recognition accuracy rate of 95.7%, showing high practicability.

In order to solve the problem of loss when the target encounters occlusion or the speed is too fast during the automatic tracking process, a target tracking algorithm based on YOLOv3 and ASMS is proposed. Firstly, the target is detected by the YOLOv3 algorithm and the initial target area to be tracked is determined. After that, the ASMS algorithm is used for tracking. The tracking effect of the target is detected and judged in real time. Repositioning is achieved by quadratic fitting positioning and the YOLOv3 algorithm when the target is lost. Finally, in order to further improve the efficiency of the algorithm, the incremental pruning method is used to compress the algorithm model. Compared with the mainstream algorithms, experimental results show that the proposed algorithm can solve the lost problem when the tracking target is occluded, improving the accuracy of target detection and tracking. It also has advantages of low computational complexity, time-consuming, and high real-time performance.

Traditional schemes for Shack-Hartmann wavefront reconstruction can be classified into zonal and modal methods. The zonal methods are good at reconstructing the local details of the wavefront, but are sensitive to the noise in the slope data. The modal methods are much more robust to the noise, but they have limited capability of recovering the local details of the wavefront. In this paper, a B-spline based fast wavefront reconstruction algorithm in which the wavefront is expanded to the linear combination of bi-variable B-spline curved surfaces is proposed. Then, a method based on successive over relaxation (SOR) algorithm is proposed to fast reconstruct the wavefront. Experimental results show that the proposed algorithm can recover the local details of the wavefront as good as the zonal methods, while is much more robust to the slope noise.

To solve the problem that the bionic compound eye system can't zoom adaptively, a zoomable bionic compound eye system based on electrowetting-on-dielectric liquid lens cambered array is proposed. The influence of the system structure on the imaging performance is analyzed, and the adaptive zoom capability of the system and the moving range of the corresponding image plane are calculated. The results show that the field of view angle increases with the increase of the curvature of the base. Compared with the non-uniform lens array, the uniform lens array can significantly reduce the defocus aberration of the system. Reducing the size of the lens unit can also decrease the defocus aberration of the edge lens. When the object distance or receiver position is changed, the defocus aberration of the system will be reduced by adjusting the focal length of the lens unit. The movable range of the system receiver is 1.9 mm~15 mm.

The property of reflective polarization on the surface of two 3D strong topological insulators was studied, obtaining the generalized necessary conditions required for the complete polarization conversion of linearly polarized light. By analyzing the reflectivity, cross reflectivity, and polarization conversion ratios of the interface of two topological insulators, we found that the model can realize the complete polarization conversion by using the existing topological insulator material, breaking through the limitation that the complete conversion requires a new small dielectric constant topological insulator material. The process can be verified by the Kerr rotation angle. Finally, we show the design method of polarization conversion devices to realize super strong angular stability. The polarization control capability of topological insulators can also be verified by Kerr effect. This work provides a theoretical basis for the application of topological insulators in polarized devices.

With the development of human-computer interaction, virtual reality, and other related fields, human posture recognition has become a hot research topic. Since the human body belongs to a non-rigid model and has time-varying characteristics, the accuracy and robustness of recognition are not ideal. Based on the KinectV2 somatosensory camera to collect skeletal information, this paper proposes a one-shot learning model matching method based on human body angle and distance characteristics. First, feature extraction is performed on the collected bone information, and the joint point vector angle and joint point displacement are calculated and a threshold is set. Secondly, the pose to be measured is matched with the template pose, and the recognition is successful if the threshold limit is met. Experimental results show that the method can detect and recognize human poses within the defined threshold in real-time, which improves the accuracy and robustness of recognition.

An algorithm (Re-CRNN) of image processing is proposed using RGB-D object recognition, which is improved based on a double stream convolutional recursive neural network, in order to improve the accuracy of object recognition. Re-CRNN combines RGB image with depth optical information, the double stream convolutional neural network (CNN) is improved based on the idea of residual learning as follows: top-level feature fusion unit is added into the network, the representation of federation feature is learning in RGB images and depth images and the high-level features are integrated in across channels of the extracted RGB images and depth images information, after that, the probability distribution was generated by Softmax. Finally, the experiment was carried out on the standard RGB-D data set. The experimental results show that the accuracy was 94.1% using Re-CRNN algorithm for the RGB-D object recognition, which was significantly improved compared with the existing image-based object recognition methods.

The principal component analysis of linear kernel and XGBoost models are introduced to design electroencephalogram (EEG) classification algorithm of four emotional states under continuous audio-visual stimulation. In order to reflect universality, the traditional power spectral density (PSD) is used as the feature of EEG signal, and the feature importance measure under the weight index is obtained with XGBoost learning. Then linear kernel principal component analysis is used to process the threshold selected features and send them to XGBoost model for recognition. According to the experimental analysis, gamma-band plays a more important role than other bands in XGBoost model recognition; in addition, for distribution on channels, the central, parietal, and right occipital regions play a more important role than other brain regions. The recognition accuracy of this algorithm is 78.4% and 92.6% respectively under the two recognition schemes of subjects all participation (SAP) and subject single dependent (SSD). Compared with other literature, this algorithm has made a great improvement. The scheme proposed is helpful to improve the recognition performance of brain-computer emotion system under audio-visual stimulation.

Aiming at the problem that most of the output energy of laser single-beam processing of two-dimensional codes is wasted and the processing efficiency is low, this paper adopts a multi-beam parallel processing method to improve the processing efficiency. The simulation studies the effect of the QR code processing filling rate, and contrast on recognition rate, and recognition time, and obtains a QR code recognizable range. Then use the femtosecond laser based on the parallel processing technology of the spatial light modulator to conduct experiments, and analyze the contrast and processing fill rate of the experimental results. The analysis results show that within the range of the QR code reading evaluation standard, the greater the processing filling rate is, the lower the recognition rate and the longer the recognition time are; similarly, the lower the contrast is, the lower the recognition rate and the longer the recognition time are. At the same time, experiments were performed on the single-beam laser processing of two-dimensional codes. The comparison of processing time and other parameters of the parallel processing and single-beam processing was analyzed, and the efficiency of the parallel processing was about 10 times higher than that of the single-beam processing.

In order to solve the problems in retinal vessel segmentation, such as blurred main vessel profile, broken micro-vessels, and missegmented optic disc boundary, a ghost convolution adaptive retinal vessel segmentation algorithm is proposed. The first algorithm uses ghost convolution to replace the common convolution in neural network, and the ghost convolution generates rich vascular feature maps to make the target feature extraction fully carried out. Secondly, the generated feature images are adaptive fusion and input to the decoding layer for classification. Adaptive fusion can capture image information at multiple scales and save details with high quality. Thirdly, in the process of accurately locating vascular pixels and solving image texture loss, a dual-pathway attention guiding structure is constructed to effectively combine the feature map at the bottom and the feature map at the top of the network to improve the accuracy of vascular segmentation. At the same time, Cross-Dice Loss function was introduced to suppress the problem of uneven positive and negative samples and reduce the segmentation error caused by the small proportion of vascular pixels. Experiments were conducted on DRIVE and STARE datasets. The accuracy was 96.56% and 97.32%, the sensitivity was 84.52% and 83.12%, and the specificity was 98.25% and 98.96%, respectively, which proves the good segmentation effect.

In the actual adaptive optics control system, the time delay causes the mismatch between the correction profile generated by the corrector and the actual wavefront distortion, which leads to correction lag error. Under the atmospheric frozen flow turbulence assumption, a wavefront distortion prediction method based on motion estimation is proposed to compensate for the time delay. The template matching algorithm is used to estimate the atmospheric turbulence motion direction, according to the wavefront restored images of the reference frame and the current frame, and then the current frame is moved to predict the next frame. The prediction method applicability is evaluated, and the influence of backtracking frames on the prediction effect is discussed by comparing the simulation data of different sampling frequencies and different transverse wind speeds. The residual error is calculated with the template matching algorithm and the least recursive squares (RLS) algorithm. The simulation results show that the method performs better when the variation tendency of wavefront restored images is obvious. Therefore, the prediction effect can be maintained better in severe conditions. Finally, the prediction method is verified by using the actual observation data of Sirius, and the algorithm still keeps the prediction effect.

Silicon photomultiplier (SiPM) is an array composed of hundreds or even thousands of single photon avalanche diode (SPAD). It has the advantages of high gain, easy integration into the array and anti-interference, and has a wide range of application prospects in LiDAR ranging. We analyzed the laser radar ranging signal and noise model, simulated the output of SiPM under sunlight, and obtained the corresponding mean value through Gaussian fitting. On the basis that the probability of being excited after SPAD received photons obeys the Poisson distribution, the calculation formula of SiPM analog output affected by background light when the background photoelectron is 0.001 /ns~0.01 /ns and the detector dead time is 5 ns~50 ns was given. Finally, we derived the expression of target detection probability. The experimental results of laser ranging detection probability under outdoor ambient light are consistent with theoretical calculations.

In order to solve the problem that SiamMask cannot adapt to the change of target appearance and the lack of use of feature information leads to rough mask generation, this paper proposes a video object segmentation algorithm based on the adaptive template update and the multi-feature fusion. First of all, the algorithm adaptively updates the template using the segmentation results of each frame; secondly, the hybrid pooling module is used to enhance the features extracted in the fourth stage of the backbone network, and the enhanced features are fused with the rough mask; finally, the feature fusion module is used to refine the rough mask stage by stage, which can effectively combine the spliced features. Experimental results show that, compared with SiamMask, the performance of the proposed algorithm is significantly improved. On the DAVIS2016 data-set, the region similarity and contour similarity of this algorithm are 0.727 and 0.696, respectively, which is 1.0% and 1.8% higher than that of the benchmark algorithm, and the speed reaches 40.2 f/s. On the DAVIS2017 data-set, the region similarity and contour similarity of this algorithm are 0.567 and 0.615, respectively, which is 2.4% and 3.0% higher than that of the benchmark algorithm, and the speed reaches 42.6 f/s.

Dispersion compensation for the data processing of the spectral domain optical coherence tomography (SD-OCT) system is an important way to improve the imaging quality of the system. A dispersion compensation method for spectral domain optical coherence tomography based on numerical polynomial fitting analysis is proposed in this paper. This method obtains the dispersion factor by fitting the phase of the interference signal and removes the dispersion mismatch terms, which can significantly improve the system axial resolution compared with non-dispersion compensation. The SD-OCT system is used to measure the axial resolution and signal-to-noise ratio (SNR) at different positions of the optical path difference, and the effectiveness and reliability of the method are verified by analyzing the axial resolution and the SNR of the system before and after the dispersion compensation technology. Finally, we found that the third-order dispersion compensation has a visible optimization effect within the imaging depth of ~1.5 mm.

Polymer membranes are attractive mirror candidate for the space large aperture lightweight optical imaging system. But there are strict requirements for mirror material because of the harsh space application environment and the high optical imaging quality requirement. The dimensional stability is one of the most important properties for optical mirror material. In this research, based on the molecular structure design, rigid molecular chain and hydrogen chain have been introduced to polyimide to improve the thermal dimensional stability. At the same time, the excellent mechanical, optical and thermal properties of the polyimide membrane have been guaranteed. The obtained optical grade polyimide has high dimensional stability, and the optical stability of the space environment and the excellent comprehensive properties are good candidates for the lightweight optical application.

Image captured in foggy weather often exhibits low contrast and poor image quality, which may have a negative impact on computer vision applications. Aiming at these problems, we propose an image dehazing algo-rithm by combining light field technology with atmospheric scattering model. Firstly, taking the advantages of cap-turing multi-view information from light field camera is used to extracting defocus cues and correspondence cues, which are used to estimating the depth information of hazy images, and use the obtained depth information to cal-culating the scene’s initial transmission. Then use scene depth information to build a new weight function, and com-bined it with 1-norm context regularization to optimizing the initial transmission map iteratively. Finally, the central perspective image of hazy light field images is dehazed using atmospheric scattering model to obtain the final de-hazed images. Experimental results on synthetic hazy images and real hazy images demonstrate that, compared to existing single image dehazing algorithms, the peak signal to noise ratio get 2 dB improvement and the structural similarity raise about 0.04. Moreover, our approach preserves more fine structural information of images and has faithful color fidelity, thus yielding a superior image dehazing result.

Vibration rejection is a key technology of telescopes with stable accuracy of μrad level. Due to the low-rate sample and large time delay of the image sensor, the conventional control loop cannot well mitigate vibrations, especially the wideband vibrations with wide range and large energy. On the concept of optimal force design, an improved wideband vibration rejection method based on Youla parameterization is proposed to mitigate vibrations for improving the closed-loop performance of telescopes. In the case that the disturbances frequency can be obtained, this method can mitigate wideband vibrations by designing an appropriate Q-filter to accommodate to the wideband vibrations. The simulation and experimental results show that the proposed method greatly improves the wideband vibration rejection ability and closed-loop performance of the system compared with the traditional proportional- integral control method. In addition, this method can be extended to many engineering projects because of its low dependence and easy implementation.

In this paper, the electrical impedance spectroscopy characteristics of polymer dispersed liquid crystal (PDLC) doped with nano-zinc oxide rods and its sensing applications are studied. Polymer dispersed liquid crystal films have the characteristics of stable structure, resistance to mechanical impact and easy preparation. By doping nano-zinc oxide rods into the material, the sensing function of polar molecules such as ethanol gas can be realized through the analysis of electrical impedance spectroscopy. In this paper, the complex impedance spectra of thin films encountering ethanol molecules are studied and analyzed through comparative experiments. In addition, the elec-trochemical equivalent circuit was established and analyzed. It was found that the film could sensitively and effec-tively realize the sensing function of the ethanol molecules. The sensitivity and response time of the sensor are fur-ther analyzed and studied. The experimental study and analysis show that nano-zinc oxide rod doped PDLC film is expected to be used as a gas sensor for detecting polarity of ethanol and other materials. It has the advantages of high sensitivity, stable structure, high repeatability, and easy fabrication.

Aiming at the application requirement of resonator fiber optic gyroscopes, a frequency tracking and lock-ing control scheme based on laser temperature and PZT control is proposed in this paper. By taking advantages ofthe large range of laser temperature tuning as well as the high precision and high dynamicity of PZT tuning, trackingof the fiber laser’s central frequency to the fiber ring resonator’s resonance frequency is realized. Typical transmis-sion resonant curve is simulated by mathematical methods. Hardware design, algorithm simulations of temperatureand PZT control scheme are carried out. The influence of control parameters on tracking stability is analyzed. Thedevelopment of laser frequency tracking systems is assembled. The high-precision and long-time tracking of laser’scentral frequency to resonance frequency is realized successfully. The locking precision is low to 4.8×10-9hour under room temperature. The locking precision is low to 9.74×10-8 over 5.5 hours under variable temperature. over oneThis work has laid an important foundation for improving the long-term performance of resonator fiber optic gyros-copes.

At present, the ground layer adaptive optical systems are using multiple laser guide stars arranged in regular polygons as reference targets to measure the effects of atmospheric turbulence. Obtaining the optimal posi-tion of laser guide stars becomes an interesting problem to analyze. This paper proposes a method to obtain the optimal position of laser guide stars by using a genetic algorithm as the optimization algorithm and a simplified geometry model of the ground layer adaptive optic system as the evaluation function. Furthermore, multi-process, Numba library, and multi-thread techniques methods are used to accelerate calculation speed. Based on these me-thods, real atmospheric turbulence profiles are used to analyze the relationship between the optimal position of laser guide stars with different numbers and the different atmospheric turbulence profiles from the same site, through an example of a ground layer adaptive optics system with 14 arcmin field of view. The results show that the optimal po-sition of laser guide stars in the same site is almost the same and their statistically optimal positions are all regular polygon. Besides, we also find that the spatial resolution of turbulence profiles has strong effects to positions of laser guide stars, showing that the more equivalent layers in the measurement results, the closer the position distribution of laser guide stars is to the regular polygon.

The main function of the laser communication large-caliber ground optical transceiver is to establish a communication link with the satellite to realize data transmission between the satellite and the ground station. The 600 mm microcrystalline primary mirror of one laser communication station is heavy, and its working angle changes constantly. In order to decrease the mirror surface error, the support system not only has a 9-pose axial support structure but also simultaneously balances the radial component of gravity of the primary mirror at its working angle by using a radial support structure. Flexible lateral support structures have large size and stress, so it is not suit for the mirror that works in a wide range of rotation. The paper researches the lateral support structure with a mercury band and central shaft, and analyses the impact of mercury band parameters on the surface error. The designed lateral structure has small size and improves the surface quality of the mirror. The measured values of PV and RMS are smaller than λ/5 and λ/37, respectively. These result shows that the designed lateral support structure reaches the design purpose and satisfies the requirements.

The field calibration of straightness is an important method to ensure the accuracy of on-line measure-ment. Based on the transceiver integrated laser five-degree-of-freedom measurement structure, the field calibration model was established aiming at the Abbe error, and the imaging error of retroreflector caused by the calibration platform. According to the calibration model and the angle measurement results of the five-degree-of-freedom measuring device, a compensation method of straightness calibration errors was proposed. Experimental results showed that the calibration coefficient error was within 0.2% when using the calibration method, and the calibration errors of straightness were effectively reduced. The calibration method made the error of calibration coefficient re-duce to less than 0.2%, and effectively improved the accuracy of straightness field calibration.

The distance between the laser optical axis and the tracking optical axis of the theodolite (axis shift) and the parallelism error of the optical axis cause the tracking position of the theodolite to be inconsistent with the laser pointing position.The analysis of the influence of off-axis and parallelism errors shows that a large amount of shift and parallelism errors will lead to inconsistencies between the laser pointing and the theodolite tracking pointing, which in turn leads to an increase in the blind zone of the laser ranging, a decrease in the operating distance and the accuracy of target positioning. A dynamic correction method for laser pointing based on bias tracking is proposed. By keeping the target always at the center of the laser beam and keeping the laser ranging position consistent with the theodolite tracking and locking position, it effectively solves the effect of laser edge energy drop on the operating distance. For a certain type of theodolite, the blind spot of the target can be reduced from 1 km to 82 m. At the same time, in view of the problem that the bias tracking algorithm needs the initial distance of the target to start the bias tracking, a one-dimensional search method for the target with unknown initial distance is proposed, which greatly improves the search efficiency of the target with unknown initial distance. The method in this paper solves the problem of the consistency between the tracking position of the theodolite and the pointing position of the laser, and greatly reduces the limitation on the shift and parallelism of the laser optical axis and the theodolite tracking optical axis.

Lyot filter is widely used in solar observation for spectra-scanning imaging. Calibration experiment at regular intervals is an important work to assure the accuracy and validity of Lyot filter. This paper comes up with a new method to conduct the Lyot filter calibration experiment on-line while traditional method requires perfect stability of environment. This method uses monochromatic imaging channel and Lyot filter scanning imaging channel simultaneously, and corrects the scanning data with monochromatic imaging data to correct the impact of environment. The instability of light source caused by disturbance of observation environment is reduced. We apply the calibration method in the high-resolution multi-wavelength solar imaging system to calibrate the Lyot filter in Hɑ (656.28) scanning imaging channel and correct the scanning data with TiO band (705 nm) observation data. The result shows that this method successfully eliminate the impact of the light instability on scanning curve of Lyot filter. The difference between the ideal center and the true center of the filter is more than 0.005 nm. The accuracy of the calibration experiment and the adaptability to environment are promoted.

In order to solve the problem that the mass and the surface figure accuracy of the space reflective mirror are often contradictory in the lightweight design process, a structural optimization design of a lightweight rectangular reflective mirror of an off-axis three-reflection optical system is performed. In this study, a lightweight structure based on the center support of SiC materials is proposed. At the same time, a multi-objective optimization method is in-troduced. With the RMS value and Mass as the optimization targets at the same time, a mirror optimal structure model is obtained with a mass of 2.32 kg. Compared with the solid mirror, the lightweight ratio is 73.8%. Then the mirror subassembly is designed and the integrated performance of it is simulated. It shows that the RMS value of the mirror reaches respectively 2.5 nm, 2.2 nm and 7.3 nm when gravity load is applied in the directions of X, Y and Z axes. Furthermore, the RMS value is 3.2 nm when the mirror subassembly is under the load condition of uniform temperature rise of 4 ℃, which is far less than the requirement of RMS≤λ/50(λ=632.8 nm). Therefore the data meets the design requirements.

The in-situ measurement of complex optical surfaces is a challenging task in precision engineering. The phase measuring deflectometry is a powerful measuring method for complex specular surfaces, and it has higher measuring efficiency, stability and dynamic range compared to interferometry. Consequently it is promising to wide-spread applications in various fields. Deflectometry is essentially a calibration problem, and the measuring accuracy is directly determined by the quality of geometrical calibration. An in-situ deflectometric measuring system is de-signed based on the single point diamond turning machine. A self-calibration method is developed to specify the relative positions of the camera and screen. Ray tracing is conducted at two positions of an auxiliary reflecting mirror, which is mounted on an air bearing spindle. The accuracy of the geometrical positions can be improved by an order of magnitude by minimizing the deviations of the traced points with respect to the true correspondences. According to the statistical properties of the deviations in reverse ray tracing, the form errors and the position errors can be separated, and the positioning error of the workpiece can be corrected accordingly. Henceforth, the nominal shape of the fabricated workpiece can be fully utilized, and the conventional one-way position-form mapping can be converted into a two-way mapping problem. As for the complex shapes, the whole surface can be covered by sub-aperture measurement. Precise localization of a local region under test is achieved by multi-position imaging, so that correct convergence of the iterative reconstruction process can be guaranteed. Several typical optical surfaces including an off-axis paraboloid mirror are measured, and the measuring accuracy of the proposed method is proved better than 150 nm RMS.

Passive hydraulic support units (PHSUs) are frequently used in the in-situ fabrication and testing (in-situ support). However, the difference in PHSUs’ stiffness will affect the mirror surface figure, especially for those thin meniscus mirrors. In order to solve this problem, the joint optimization method of layout, stiffness and active correction is studied. Firstly, for the difference of PHUS' stiffness, a hierarchical layout optimization method for support stiffness and support position is proposed to obtain the initial optimization solution of the support system. Then, the mode calibration method and the least square method is used for active correction of support system to obtain the final optimized solution of the mirror surface figure. Finally, the effectiveness of the method is verified by a numerical simulation experiment with specific cases. The results show that, for 4 m thin meniscus mirror, after layout optimization, with the hydraulic passive support system, the root mean square (RMS) of the mirror surface errors of 60 point axial support system is reduced from 150.6 nm to 32.9 nm, and the RMS value of the mirror surface errors of 78 point axial support system is reduced from 45.2 nm to 22.6 nm. The optimization effect is remarkable. After active correction, the RMS value of the mirror surface errors of 60 point axial support system is 14.6 nm, and it is 6.9 nm for 78 point axial support system. The requirement of the RMS value of the mirror surface error is less than λ/40 (λ=632.8 nm). The support systems meet the requirement. Finally, the 60 point axial support system was selected. Through the joint optimization of layout, stiffness and active correction for supporting points, it can greatly increase the applicability, flexibility and reduce the difficulty of implementation for the in-situ support system.

Stressed polishing technology transforms aspheric fabrication into spherical fabrication by applying pre-determined loads on the surface of the mirror. The key to achieve high precision of stressed polishing is to test the surface deformation with high precision. Stereoscopic phase measuring deflectometry was used to test the surface topography and the deformation of stressed mirror. After obtained unwrapped phase distribution, and combined with normal consistency constraint and gradient integral algorithm, the height distribution was finally obtained. Composi-tion of systematic errors were simulated. Also, the errors were calibrated and removed by N-step averaging method in this system, which improved the measuring precision. In this paper, the surface topography and the deformation of a stressed mirror with a diameter of 320 mm, spherical radius of 5200 mm were measured. The measuring results were consistent with the corresponding result of CMM and finite element simulation, indicating that this proposed method is on the level of micron in terms of accuracy and more suitable for the test of stressed mirror compared with interferometer and CMM.

The fabrication of optical elements with microstructural arrays has attracted more and more attention. Single-point diamond flying cutting technology has been gradually applied to the fabrication of microstructures with the advantages of high efficiency, low cost and high machining accuracy. This paper mainly studies the influence of repeated positioning errors of machine tools and errors introduced by cyclic machining on micro-structure turning effect when flying cutter turning micro-pyramid structure, analyses the conditions of secondary groove generation in V-groove turning, studies the methods of restraining secondary groove generation, and finally verifies through expe-riments that the generation of secondary groove can be restrained by controlling the turning depth greater than the minimum turning depth.

Side-scan sonar (SSS) is an electronic device that utilizes the propagation characteristics of sound waves under water to complete underwater detection. Because the SSS produces images and maps according to the in-tensity of acoustic echo, speckle noise will be inevitably involved. A speckle denoising method based on block-matching and 3D filtering (BM3D) is proposed to filter the multiplicative speckle noise in SSS images. First, the SSS image is transformed by power and logarithm. The wavelet transform is used to estimate the general noisy level of the polluted image. Second, the parameters of the BM3D algorithm are updated according to the noise estimation results of each local patch. At last, after comparing the general noise estimation and the local noise estimation, the proposed algorithm chooses the best estimation to filter every patch separately to solve the problem that the noise is not evenly distributed. The experimental results show that the improved BM3D algorithm can effectively reduce the speckle noise in SSS images and obtain good visual effects. The Equivalent Number of Looks of the proposed al-gorithm is at least 6.83% higher, the Speckle Suppression Index is lower than traditional algorithm, and the Speckle Suppression and Mean Preservation Index is reduced by at least 3.30%. This method is mainly used for sonar image noise reduction, and has certain practical values for ultrasonic, radar or OCT images polluted by speckle noise.

To solve the unevenness of distributions of optical illuminance and power in visible light communication system, a light source layout based on multi-population genetic algorithm is proposed. Taking 15 LED lamps as an example, the position coordinates were optimized under the fitness function related to variance of received power through the co-evolution of multi-populations. The simulation results on Matlab R2016a showed that, after being op-timized, the distribution of power was evener intuitively, the variance of power reached 1.5744 dBm, the illuminance fell in a range between 889 lx and 1009 lx and the uniformity ratio of illuminance was 91.73%, all of which were better than those of the layout optimized by traditional genetic algorithm and the rectangular layout optimized by mul-ti-population genetic algorithm. This experiment provides a feasible solution for optimizing the visible light commu-nication system so that users can have a more comfortable communication trip in this system.

With the problem of difficulty that a single filter to adapt to various complex changes in the tracking process, an adaptive multi-filter target tracking algorithm based on the efficient convolution operators for tracking is proposed. Spatial-temporal regularized filter, the consistency check filter and the correlation filter in the efficient convolution operator tracker, convolve with target features respectively, which obtains three detection scores. The training method of spatial-temporal regularized filter is to introduce temporal regularization into loss function. The consistency check filter is a filter that uses current filter to track the target of previous several frames and updates only when the error of forward and backward position is less than the threshold. Target position is estimated by the best filter detection score with the peak-to-side ratio is maximum. The improved algorithm is tested with the OTB-2015 dataset and UAV123 dataset. The experimental results show that the proposed algorithm can better adapt to the complex environment in tracking process, which has high precision and robustness.

The observation and recognition of sunspots is an important task of solar physics. By observing and analyzing sunspots, solar physicists are able to analyze and predict solar activities with higher accuracy. With the continuous progress of observation instruments, solar full-disk image data amount is also on a rapid growth. In order to recognize and label sunspots quickly and accurately, a two-layer sunspot recognition model is proposed in this paper. The first layer model is based on deep learning model YOLO. In order to enhance the ability of YOLO to recognize small sunspots, the parameters of YOLO are optimized by using the k-means algorithm based on inter-section-over-union. The final YOLO model can identify most large sunspots and sunspot groups, with only a few isolated small sunspots being unidentified. For the purpose of further improving recognition rate of small sunspots, the second layer model applies AGAST (adaptive and generic accelerated segment test) feature detection algorithm to specifically identify the missing small sunspots. The experimental results on SDO/HMI sunspot data set show that all kinds of sunspots can be recognized effectively with high recognition accuracy by using the model proposed in this paper, thus realizing the real-time sunspot detection task.

A 3D projection system based on complementary multiband bandpass filter (CMBF) is proposed in this paper, which enables viewers to gain 3D experience through special glasses. Different from the time-multiplex or the spatial-multiplex system, it is a spectrum-multiplex system using pairs of CMBFs. The three pairs of complementary bandpass of a pair of CMBFs can be designed to cover the three spectrum ranges of RGB individually and in each pair the two bandpass nearly do not overlap. In this paper, a 3D projection system is built from two ordinary projec-tors and its spectrum, brightness and crosstalk have been measured. The average crosstalk is 3%, meeting the ba-sic requirement of crosstalk in 3D display which is less than 10%.

At present, several mainstream algorithms using color name (CN) all adopt principal component analysis (PCA) to process the feature. However, PCA assumes that the noise of input data must obey Gaussian distribution, which is a conspicuous defect. Aim to address this problem, in this paper, we take robust principal component analysis (Robust PCA) to process CN features. The method projects the original RGB color space to a robust color space–CN space, which means that the input image is stratified to 11 layers according to color name. Then, it processes the CN features by the Robust PCA, so that the mapped image information is concentrated on a few layers, retaining a great quantity of image information and filting out noise. The processed feature is used for Color-tracking frame at the standard benchmark OTB100, and we set up different layers to compare the performance differences of the algorithm. The experimental results show that the success rate increases by 1.0% and the accu-racy increases by 0.9% at OTB100. The result illustrates that the Robust PCA method can better bring color name feature superiority into full play and improve the performance of the algorithm effectively.

A preternatural and extremely thin metasurface with weak asymmetric unit structure is presented here to demonstrate extraordinary strong chirality. The unit cell of metasurface is composed of a double layer of elliptical metal patches with a certain twisted angle and a medium sandwiched between them. When the twisted angle equals to 80°, optical activity can be realized in this metasurface. At the resonant frequency 11.89 GHz, the incident linearly polarized wave is converted into its cross-polarization wave with the transmittance rate higher than 94%. The light weight and miniaturization of this metasurface provide a reliable approach for polarization manipulation. If extended to light waveband, the metasurface may have potentials in biological applications such as detection of weak chiral molecules, etc.

Blind image deconvolution is one method of restoring both kernel and real sharp image only from de-graded images, due to its illness, image priors are necessarily applied to constrain the solution. Given the fact that traditional image gradient l2 and l1 norm priors cannot describe the gradient distribution of natural images, in this paper, the image sparse prior is applied to the restoration of single-frame atmospheric turbulence degraded images. Kernel estimation is performed first, followed by non-blind restoration and the split Bregman algorithm is used to solve the non-convex cost function. Simulation results show that compared with total variation priori, sparse priori is better at kernel estimation, producing sharp edges and removal of ringing, etc., which reducing the kernel estimation error and improving restoration quality. Finally, the real turbulence-degraded images are restored.

Orbital angular momentum (OAM) beam with helical phase distribution has demonstrated important ap-plications in information optics, optical trapping, and optical manipulation. In this paper, we designed a planar optical device which can generate a periodic array of focused orbital angular momentum beams. Based on detour phase encoding, the phase distribution calculated by fractional Talbot effect is implemented on this planar optical device. The property of this optical device with periodic square and hexagonal structures is simulated by finite difference time domain (FDTD) respectively. The optical device with explicit advantages of being easy to fabricate, splice, dup-licate, and integrate can efficiently prop up the generation of high-quality large-area array-type OAM beams.

Based on the Rytov approximation theory, we analyze the cross-spectral density of Hankel-Bessel (HB) beams in anisotropic ocean turbulence. In this paper, we study the orbital angular momentum (OAM) mode detection probability, the crosstalk probability and the spiral phase spectrum of the HB beam, and establish the OAM mode detection probability model in anisotropic ocean turbulence. The results show that the detection probability of the emission mode is decreased and the spiral phase spectrum is expanded due to the ocean turbulence. Furthermore, with the increase of anisotropy factor, the influence of ocean turbulence on the detection probability of HB beam becomes smaller. Meanwhile, with the increase of the temperature variance dissipation rate and the equilibrium parameter, and the decrease of the dynamic energy dissipation rate, the influence of ocean turbulence on the orbital angular momentum transmission is increased.

Au film is mainly used to prepare the metal structure of the terahertz (THz) microstructure. When the metal structure is fixed, it is difficult to control the terahertz wave by using the properties of Au film. In this paper, the tera-hertz microstructure based on the soft magnetic FeNHf film with the high permeability is designed and fabricated on the high resistivity silicon substrate. The magnetization direction of soft magnetic film is controlled by the external magnetic field H. The THz transmission characteristics and electromagnetic resonance mode of the microstructure under the control of H in split triangular structure are systematically studied. The soft magnetic FeNHf film has the characteristic of magnetic anisotropy. Therefore, the direction of the magnetization M in FeNHf film can be controlled by the external magnetic field H to be perpendicular and parallel to the magnetic field of THz wave, respectively. The THz time domain spectroscopy system is used to test the terahertz transmission characteristic of the microstructure. The finite difference time domain method is used to analyze the THz electromagnetic field distribution and modula-tion mechanism based on the microstructure of the FeNHf film. The experimental results show that the resonance frequency of the split triangular THz microstructure can be modulated under magnetic field. At the frequency of 1.3 THz, the tunability and modulation depth are about 5.7% and 15%, respectively.

To overcome the disadvantages of narrow frequency band and low transmittance for traditional na-no-antenna, a nano-antenna structure based on cross-slots fractal was designed. The extraordinary optical trans-mission characteristics of the cross-slots fractal nano-antenna and the differences between the cross-slots fractal nano-antenna and the uniform cross-slots nano-antenna were analyzed by the finite difference time domain method. Meanwhile, the influence of physical parameters on the extraordinary optical transmission characteristics of the cross-slots fractal nano-antenna and the relationship of transmission spectrum of the nano-antenna between the fractal size and the non-fractal size were discussed. The results show that the fractal cross-slots structure is more miniaturized, and realizes extraordinary optical transmission and full 2π phase control of transmission beam, and the transmittance is higher than the uniform cross-slots structure, the full width at half maximum (FWHM) is wider, and the highest transmittance is up to 99.51%. By adjusting the physical parameters, the transmission spectrum exhibits red-shift or blue-shift characteristics, achieving controllability of the transmission spectrum. When h=50 nm, the full width at half maximum is about 356 nm, and the transmittance is still as high as 95.66%, which is generally higher than traditional structures, and the peak transmittance is still greater than 74% at large incident angles (70 degrees). In short, the cross-slots fractal nano-antenna has the characteristics of wide frequency, controllable and adjustable, and more miniaturized structure compared with other nano-antenna structures, and realizes extraordinary optical transmission.

Aiming at the effect of pixel defects on the display of electrowetting electronic paper, an automatic thre-shold detection method based on Otsu is proposed to detect defects. Otsu is a commonly used automatic threshold method that gives satisfactory results when the image histogram is bimodal. However, the electrowetting defect image histogram is usually a single peak, and Otsu method fails. Electrowetting differs from the background contrast due to the filling inks of different colors, making segmentation more difficult. In this paper, the weighting coefficient is introduced before the target variance, and the weight decreases as the cumulative probability of defects increases. The weight keeps a large value before the threshold crosses the peak, and the weight decreases after the peak, ensuring that the threshold is always to the left of the peak in the case of a single peak. The experimental results show that the proposed method can effectively segment the electrowetting defect region, especially in the electro-wetting defect image with lower contrast ratio. The method is closer to 0 compared to the ME value of Otsu, VE, WOV and entropy weighting methods. The proposed method has a better segmentation effect.

For the digital micromirror device (DMD) lithography equipment, due to the exposed images joint errors which caused by mechanical loading errors, problems such as misalignment and overlap of the exposed images may arise. In order to eliminate the exposure error of DMD during large-area exposure, the error correction method was studied. Firstly, the exposure error was got by measuring the exposed substrate with a microscope. Then, an error model was established based on the known exposure error. Finally, an error correction based on motion com-pensation for DMD lithography system was proposed based on the error model. This method is different from the existing error correction method. The experimental results show that during the micron image exposure process, the exposure error is reduced by more than 80%, and the DMD exposure center offset distance is reduced from 175 μm to 21 μm. The stitching accuracy of the exposed image is improved effectively, which meets the requirements for high quality and high precision of large-area exposure images.

The reconstruction of wavefront from single far-field image data has unique advantages in simplicity of structure. However, the traditional wavefront reconstruction algorithm has multiple solutions based on single far-field image, its iterative process easily falls into stagnation. In this paper, based on the analysis of the multi-solution problem of single-frame phase retrieval method, a wavefront reconstruction method based on Walsh function two-dimensional discrete phase modulation is proposed. This method can effectively break the symmetry of near-field wavefront and overcome problem of multiple solutions. The simulation results show that the method can accurately reconstruct wavefront aberration with only one far-field image.

In this paper, a kind of temperature sensor which can detect a small-area heat source with high sensitivity is designed by using the property of different thermal expansion coefficients of materials. The temperature sensitive element of the sensor is a silicon nitride cantilever beam which is coated with metal on its upper surface. Due to the difference of thermal expansion coefficients between the metal and silicon nitride, the cantilever beam will bend in the direction of rapid change of the temperature gradient, and the bending amount will be positively correlated with the temperature when the ambient temperature of the cantilever beam changes. In the experiment, the bending amount of the beam is measured by the optical lever, and the relationship between the temperature and the output voltage of the detector is established by calibration. The results show that the sensitivity of the sensor can reach 4.86 mV/℃ and the temperature resolution can reach 0.04 ℃. In order to verify the applicability of the sensor for measuring the small-area heat source, the heat generated by heat sources of different areas is measured depending on the calorific property of NaYF4 under laser excitation. The results show that it still can be measured even the heatin area is only 0.07 mm2 and the accurate measurement for temperature of the small-area heat source can be realized.

Quantum dot materials have the characteristics of narrow luminescence spectrum, adjustable luminescence wavelength and high fluorescence quantum yield. The quantum dot LEDs have more potential in improving color gamut. In this paper, a method of white light generation by blue LED excited CdSe red and green quantum dots is introduced. The ratio of red and green quantum dots to glue was (1∶60) and (1∶10), and the test range of glue content was 1.4 μL～2.2 μL and 3.0 μL～5.0 μL. The samples were prepared by traditional method and layered structure. The absorption and conversion ratio of blue light in the range of glue amount were tested. The function relationship between glue amount and absorption and conversion was obtained by Matlab fitting. When taking the dots (0.34, 0.3) in white light region formed in the above test glue amount range, the red and green quantum dots were calculated with glue amount of 1.9 μL and 4.55 μL, and the corresponding theoretical spectrum was established according to the spectral calculation formula. According to the above glue amount, the verification samples were made and tested the color coordinates (0.3409, 0.2992) and the homologous spectra were basically coincident with the theoretical spectra.

Catenary nanostructures enable continuous phase control. However, the ordinary catenary nanostructure has narrow width at both ends and is not easy to be fabricated. On the other side, it was difficult to build complex model directly in simulation software CST, and the simulation process was complicated in the past. The equal-width catenary slit is proposed to replace the normal catenary slit. And the equal-width catenary-type metasurface has been designed to generate Bessel beam, which provides a new idea for the design of two-dimensional optical de-vices. In the process of modeling and simulation, CST is combined with Matlab for co-simulation, and all operations, such as modeling, simulations, and parameter modification, are completed directly in Matlab. This method can be used to design complex structures, and more ideal simulation results can be obtained combined with the numerical optimization ability of Matlab.

In order to reduce the influence of noise on the output signal of FOG, a de-noising algorithm of FOG based on modified ensemble empirical mode decomposition (MEEMD) and forward linear prediction (FLP) is proposed. Firstly, the concept of permutation entropy is introduced, and the FOG signal is decomposed and reconstructed by using MEEMD. Secondly, the low-order IMF terms of the mixed noise after decomposition is filtered and de-noised by the FLP algorithm. Finally, the signal processed by the MEEMD-FLP is reconstructed to get the result. The static test of a FOG is carried out. The experimental results show that compared with the original FOG signal, the RMSE after de-noising is reduced by 76.77%, and the standard deviation is reduced by 76.76%. It can effectively reduce the influence of noise on the FOG output signal and has higher de-noising accuracy.

In recent years, terahertz imaging has attracted great attention due to its advantages including penetrability and nondestructive property. The field imaging technology within the terahertz range is expected to enhance the terahertz image quality and improve its application effect. In this paper, an experiment on the data acquisition and digital refocusing of the terahertz light field was conducted. Firstly, the basic principle, system structure, and the method of reconstructing light field imaging were analyzed. Secondly, the terahertz focal plane array camera was used to collect the data about light field and digital refocusing was used to get the computed imaging. Finally, the reconstructed image was enhanced to obtain higher depth resolution, angle resolution, and object contour resolution. Experimental results showed the feasibility and ability of terahertz light field imaging to improve image quality and enrich retrieval effects.

In this paper, the theoretical model of ununiform plasma sheath is established based on scattering matrix method and the transmission characteristics of 0.1 THz～10 THz wave are simulated. A kind of plasma jet is produced in laboratory environment according to the principle of dielectric barrier discharge. Then the measurement of transmission spectrum of terahertz time-domain spectroscopy (THz-TDS), broadband terahertz source, and the terahertz wave reflective imaging of target under plasma shelter are carried out, respectively. Both theory and experiment results show that terahertz wave has good penetration in plasma, which provides a new way for communication and radar detection in blackout area.

In order to analyze the performance of vehicle positioning using LED traffic lights in foggy environment, the influences of receiving angle, road width, and signal-to-noise ratio (SNR) at the receiver in foggy environment on vehicle positioning range are discussed. The simulation results show that the optimal signal reception angle is 25°; when the vehicle is within 20 m of the LED traffic light, the road width has a greater impact on the received power of the signal; near the traffic light, the positioning distance of the vehicle on the second lane is 2.2 m shorter than that of the vehicle in the first lane; the SNR of the receiver at night is better than during the day, and the positioning range at night is greater than during the day; compared with sunny weather, the SNR in the fog days decrease significantly, which will greatly affect the positioning range of the vehicle, so to ensure safe driving, vehicles need more braking time when driving in foggy days.

A cylindrical image mosaic method based on fast camera calibration in multi-scene is proposed to solve the problems of scene limitation and complex calibration process in image mosaic using camera calibration parameter. Firstly, the accurate corner feature of checkerboard calibration board is used to make it in the overlapping field of view of two adjacent images. Then, the image sequence is pre-processed by corner extraction, precision and matching, so that the registration parameters between the images to be stitched can be solved accurately and quickly. After that, the cylindrical projection is used to maintain the visual consistency of the images, and the multi-band fusion is used to retain the details of the images. Subsequently, the images are stitched using registration parameters obtained by calibration. Finally, the whole system is built on a low-power embedded platform to accomplish fast calibration and mosaic process based on calibration parameters in multi-scene. The experiment results show that the proposed method can accomplish camera calibration quickly and accurately in indoor and tunnel scenarios, and the image mosaic process is time-consuming. Meanwhile, it can ensure better stitching accuracy and imaging effect, and has strong robustness.

Transient measurements of high-speed airflow field are needed in the measurement of boundary layer. Digital interferometry can measure flow field quantitatively to obtain density information, which is very necessary in flow field measurement. In this paper, a common-path shearing interferometry method is introduced. It is insensitive to vibration and does not need a reference plane. It is suitable for flow field measurement. A fast algorithm based on spatial phase modulation, coupled with a pulse laser and a synchronous control system, is used to measure the disturbance density field quantitatively in real time. The acquisition resolution of the system is 200 pixels × 200 pixels, and the acquisition frequency can reach more than 1000 frames per second. The wavefront reconstruction method of the system has been simulated by computer, and the detection result is better than 1/20λ. The experimental results in a 0.6 m wind tunnel show that the system can restrain the vibration interference and distinguish the disturbance signal and the vibration noise remarkably. It has good application prospects.

Starting from the expression of Laguerre-Gaussian vortex beam and based on Rayleigh diffraction theory, the variation of rotating coherence function of vortex beam propagating in atmospheric turbulence is studied. The crosstalk between the angular momentum of each orbital angular momentum when the vortex beam propagates in atmospheric turbulence is summarized. The topological charge detection probability is used to describe the crosstalk law, and the analytical expression of the topological charge detection probability is derived. The distribution of topological charge number of vortex beam passing through turbulence is studied, and the results are compared with the numerical simulation results of vortex beam passing through atmospheric random phase screen. The relationship between the detection probability of the theoretical and simulated topological charge numbers with the turbulence intensity and the topological charge number of the initial vortex beam is compared, and the correctness of the analytical expression of the topological charge number detection probability is verified. Through this expression, the interaction between atmospheric turbulence and vortex beam can be further studied, which can affect the essence of angular momentum scattering of vortex beam, and the suitable topological charge number interval can be selected for the space optical communication of vortex beam. It also provides a theoretical basis for selecting the appropriate beam waist size under different turbulence intensities to reduce the bit error rate (BER) caused by crosstalk.

In order to obtain better super-resolution reconstruction results of depth images, this paper constructs a multi-scale color image guidance depth image super-resolution reconstruction convolutional neural network. In this paper, the multi-scale fusion method is used to realize the guidance of high resolution (HR) color image features to low resolution (LR) depth image features, which is beneficial to the restoration of image details. In the process of extracting features from LR depth images, a multiple receptive field residual block (MRFRB) is constructed to extract and fuse the features of different receptive fields, and then connect and fuse the features of each MRFRB output to obtain global fusion features. Finally, the HR depth image is obtained through sub-pixel convolution layer and global fusion features. The experimental results show that the super-resolution image obtained by this method alleviates the edge distortion and artifact problems, and has better visual effects.

The geometry parameters of optical fiber affect the optical transmission and mechanical properties, which are the important indexes to measure the quality of fiber. Near-field light distribution method is recommended in GB15972.20-2008 for the measurement of geometry parameters. In order to distinguish the boundary between fiber core and cladding, the method needs to illuminate the fiber. The end face of the fiber core is a bright spot with unclear edge, so the true edge of the core and cladding cannot be accurately judged. In this paper, the distribution of mode field in optical fiber is analyzed. Theoretically, the solution of electromagnetic vector of mode field satisfies Bessel function, but Gaussian function can also be used under approximate conditions. Therefore, Gaussian function is used to fit the distribution of the fiber core in this paper, and the real edge of the fiber core and cladding can be obtained from the Gaussian function after fitting. This method is a further improvement on the measurement method of GB15972.20-2008. The experimental results show that when the cutting effect of the fiber is not good or the imaging quality is poor, the Gaussian function method fitting with mode distribution can still ensure the repeatability of the measurement and the stability of the measured data.

Aiming to the problems of traditional active electromagnetic transformer such as easy magnetic saturation, poor stability and anti-interference ability, and limited installation, etc., this paper designs the optical fiber current transformer to measure current by rotation angle based on Faraday magneto-optic effect; HB Spun optical fiber is used as sensing element without saturation and can be used for high current measurement. The designed transformer uses the optical reciprocity loop to eliminate the interference of temperature and optical fiber defect on the measurement of optical rotation angle, and uses reflector to enlarge the optical rotation angle to four times, which can realize accurate measurement of small current; Sensing element uses flexible fiber ring with shape variability characteristic, which helps for measurement of current in complex space. The paper compares flexible fiber ring with different loops to standard current transformer, the results show that optical reciprocity loop can eliminate the interference of temperature on the current measurement and the accuracy of all-fiber current transformer is 0.5 in the range of -5 ℃～70 ℃, which can realize the accurate measurement of small current.

In the adaptive optics system, the piezoelectric steering mirror(tip/tilt mirror, TTM) is usually used to correct the wavefront aberration caused by atmospheric turbulence in real time. However, the response of the piezoelectric tilting mirror has large nonlinear hysteresis effect, which greatly reduces the precision of the tilting mirror in place, affects the stability of the system, and restricts the bandwidth of the skew correction system. Therefore, the hysteresis phenomenon needs to be modeled and compensated by the established model. In this paper, hysteresis operator is introduced and using Bayesian regularization training algorithm to train BP (back propagation) neural network to construct hysteresis model of piezoelectric steering mirror. Then experimental study was conducted on a piezoelectric steering mirror developed by Institute of Optics and Electronics, Chinese Academy of Sciences. The final experimental results show that the hysteresis model of piezoelectric steering mirror constructed by BP neural network has more accurate identification capability, the hysteresis size in the X direction decreased from 6.5% to 1.3% and that in the Y direction decreased from 7.1% to 1.6%.

Study of the time characteristics of photomultiplier tube with ultra-fast time characteristics has an impor-tant guiding role for the further development of ultra-fast time response microchannel plate photomultiplier tube (FPMT). Based on the VME test system in high-energy physics and picosecond laser with single-photon pulse mode,this manuscript designs a device to test the FPMT with 25 ps system error. The time characteristics of various FPMTs were tested by optimizing the FPMT signal readout anode, the voltage divider structure and voltage division ratio. The intrinsic time lower limit value of the FPMT in the non-single-photon working mode, is proposed to compare and analyze the time resolution of different FPMTs in different working states. After completing various optimized readout anode structural design for the FPMTs, it can be find that the best FPMT prototype in our Lab has the best intrinsic time resolution lower limit of 30 ps.

In order to solve the problem that the displacement accuracy of linear displacement mechanism is toohigh in traditional white light interferometry, this paper proposes a full-field heterodyne white light interferometry. Thetechnology mainly uses the white light interference signal with difference frequency as the light source to realize thehigh-precision detection of the coherent peak position under the conditions of large push step and low push precision.In this paper, the mathematical model of white light heterodyne interference is established firstly, and then the overallsystem design scheme is proposed according to the light intensity signal characteristics provided by the mathemat-ical model. Then the feasibility of the measurement scheme is verified by experiments. At the end, theoretical anal-ysis and data comparison are carried out for the influence of various errors on the calculation accuracy of the algo-rithm. The results of error analysis show that the white-light heterodyne interferometry technology provides highermeasurement accuracy and better anti-interference performance, effectively reducing the strict dependence of tradi-tional white light interferometry on the accuracy of linear displacement mechanism, and is an optical free-form sur-face detection technology. More solutions are available.

The development direction of metal material laser processing is to achieve small roughness, less heat-affected zone and high depth-diameter ratio. Recently, a kind of water-conducting laser processing technology based on laser water-jet coupling technology has been developed. The basic principle of water-conducting laser processing technology and its advantages over traditional laser processing methods are expounded. Based on the principle of laser water-jet coupling technology, a set of water-conducting laser processing equipment is constructed. The experiments of water-conducting laser processing for various metal materials are carried out. The surfaces of work piece are observed and analyzed by Leica DVM6 digital microscope. The edges of blind holes in two kinds of metal materials are regular and smooth, the edges of grooves are straight and without burrs, and there is no heat-affected zone in both materials. The results of experiments show that water-conducting laser processing tech-nology on metal precision machining is practical and has important application value.

Aiming at the motion errors of the linear stage, a measurement method for the determination of three-degree-of-freedom (3-DOF) error motions is proposed based on non-diffracting Moiré fringes produced by computer-generated holograms (CGHs). A liquid crystal spatial light modulator (SLM) is used to generate non-diffracting beams, and two non-diffracting beams form Moiré fringes. A 3-DOF measuring optical path of non-diffracting Moiré fringes is designed. Meanwhile, a 3-DOF mathematical model of motion errors is established, and three kinds of motion errors (yaw angle, roll angle and pitch angle) are separated by geometric analysis method. A rotary table is used to simulate the 3-DOF motion errors on different conditions. The NDB and non-diffracting Moiré fringe patterns are obtained by CCD1 and CCD2 respectively. Experimental results show that the motion errors calculated by the positions of the central points agree well with the theoretical value with the error less than 0.0104°, which can verify the feasibility and correctness of the 3-DOF measurement system for non-diffracting Moiré fringes.

In order to measure the transmission wavefront of laser rods and to improve the edge diffraction effect of small-aperture laser rods measured by Tayman or Fizeau interferometer, a variable-inclination Mach-Zehnder inter-ferometer was proposed. The incident angle was changed by adjusting the tilting attitude of the phase shifting ref-lector, then the optical path difference was changed that the phase shift was introduced to the coherent light and the phase shifting interferometry was realized. The transmission wavefront of a laser rod (Nd:YAG) with the diameter of 6 mm and the length of 60 mm was measured by this interferometer, the peak-valley (PV) and root mean square (RMS) of the wavefront were 0.391λ and 0.056λ. The same laser rod was measured by ZYGO GPI XP interferometer, the peak-valley (PV) and root mean square (RMS) were 0.370λ and 0.064λ. The comparison results show that the interferometer can achieve high-precision detection of transmission wavefront of laser robs. The variable-inclination Mach-Zehnder interferometer has high phase-shifting precision and wide phase-shifting range, and the beam in the system can pass through the laser rod only once, which can suppress the multi-beam interference and improve the edge diffraction effect of the small-aperture laser rods.

Aiming at the complexity of the traditional gauge detection method, high requirements for the installation and large amount of data analysis, a gauge measurement method based on the relative transverse movement of wheel and rail is designed in this paper. Two sets of laser source and camera combinations are used to dynamically collect the image of the inner straight line part of the rail head in the method. According to the rail parameters, the Hough detection and perspective transformation are used to rectify the image from the same acquisition distance. Compared to datum moment, the variation of vertical displacement of the center point of the laser is computed. And through the geometrical relation of the variation previously described and the lateral relative displacements of the two wheels, the relative transverse displacement of the two wheels is calculated. The relative initial time gauge change is gained by the difference, which realizes the indirect measurement of the track gauge. The experimental results show that the method has the characteristics of simple hardware structure, small data calculation, high measurement precision, and can realize non-contact dynamic measurement of gauge parameters.

Sweep source optical coherence tomographic angiography (SS-OCTA) is a kind of angiography technol-ogies based on split spectrum amplitude deeorrelation angiography (SSADA). It has a great prospect in the early diagnosis of tumors and other diseases. In this paper, skin structure and angiography of melanoma C57BL6 mice were collected on the basis of the SS-OCTA imaging platform with an imaging field of 5.12 mm×5.12 mm and a standard image maximum signal-to-noise ratio of 34.3 dB. The results show that the SS-OCTA system is superior to the structural imaging in early diagnosis of dermatological diseases.

Aiming at the lowoptical efficiencyof Fresnel lens, a high-efficiency non-imaging concentrated optical(NICO) system composed of an aspheric lens and a trumpet lens was designed. The aspheric lens was optimizedinsequential mode of Zemax, and the geometric radius of its image spot was reduced from 42 mm to 1.7 mm by mi-nimizing the spherical aberration. The aspheric lens and trumpet lens were modeled and optimized in non-sequential mode of Zemax, and the NICO system with 87% optical efficiency and 0.9° received angle was achieved by Monte Carlo ray tracing analysis. Finally, the packaging and testing of the high concentrated photovoltaic (HCPV) module were completed based on samples of an aspheric lens array and 48 trumpet lenses. The test results showed that the photoelectric conversion efficiency of the module reached 30.03%, which was significantly improved compared with the HCPV module composed of the Fresnel lens.

In view of the problem that 3D-CNN can better extract the spatio-temporalfeatures in video, but it requiresa high amount of computation and memory, this paper designs an efficient 3D convolutional block to replace the 3×3×3 convolutional layer with a high amount of computation, and then proposes a 3D-efficient dense residual networks (3D-EDRNs) integrating 3D convolutional blocks for human action recognition. The efficient 3D convolu-tional block is composed of 1×3×3 convolutional layers for obtaining spatial features of video and 3×1×1 convolu-tional layers for obtaining temporal features of video. Efficient 3D convolutional blocks are combined in multiple lo-cations of dense residual network, which not only takes advantage of the advantages of easy optimization of residual blocks and feature reuse of dense connected network, but also can shorten the training time and improve the effi-ciency and performance of spatial-temporal feature extraction of the network. In the classical data set UCF101, HMDB51 and the dynamic multi-view complicated 3D database of human activity (DMV action3D), it is verified that the 3D-EDRNs combined with 3D convolutional block can significantly reduce the complexity of the model, effec-tively improve the classification performance of the network, and have the advantages of less computational re-source demand, small number of parameters and short training time.

The planar compound eye imaging system uses multiple sub-apertures to image the scene. Due to the constraint of the imaging sub-aperture size and spatial sampling rate of the image sensor, the image quality of each sub-aperture is low. How to fuse multiple sub-aperture images for a high-resolution image is an urgent problem. Multi-image super-resolution theory uses multiple images with complementary information to reconstruct high spatial resolution image. However, existing theories usually adopt the oversimplified motion model which is not suitable for planar compound eye imaging. If the existing multi-image super-resolution theory is directly applied to the resolution enhancement of the planar compound eye, the inaccurate motion estimation will reduce the performance of image resolution enhancement. In order to solve these problems, the motion model of the multi-image super-resolution is improved in the variational Bayesian framework, and the derived joint estimation algorithm is used to enhance the resolution of the planar compound eye. The correctness and effectiveness of the proposed method is verified by the simulation data experiments and the real compound eye data experiments.

In order to adjust the position of the secondary mirror of Cassegrain telescope with large field of view, a computer aided adjustment method of two-step sensitivity matrix model was proposed. Based on the analysis of the shortcomings of the traditional sensitivity matrix method of the two order model, a fine tuning step was added based on the characteristics of the sensitivity matrix and the traditional sensitivity matrix method was improved. For the Cassegrain system, the relationship between the Zernike coefficients and the misalignment was analyzed, and the calibration simulation of Cassegrain system with 300 mm aperture and 0.6° field of view was carried out. The simu-lation results show that after correction by traditional sensitivity matrix method, the mean values of offset along x, y, z axes and tilt around x,y axes are -0.0684 mm, -0.0892 mm, 0.0015 mm, 0.0498° and -0.0444°, respectively, and the full field wavefront aberration RMS is less than 0.1λ(λ=632.8 nm). After correction by two step sensitivity matrix correction method, the mean values are -0.0018 mm, -0.0012 mm, 0.0002 mm, 0.0008° and -0.0012°, respectively, and the full field wavefront aberration RMS is less than 0.03λ, which is obviously superior to the traditional sensitivity matrix method.