Aiming at the problems of complex structures and difficult debugging in the current design based on prism-grating combination, we established a mathematical model of dispersion based on prism-grating combination and proposed an optimization method of only changing the relative positions of the prism and grating to increase the spectral dispersion linearity. As a result, we obtained the parameters of the prism-grating combination structure with good linearity. Furthermore, we used optical standard elements to simulate and experimentally verify the correctness of the model, and the results showed that our system had the advantages of 15 cm length, good spectral linearity, and simple structure in the band of 400-1000 nm. Finally, the experiments demonstrated that the overall resolution was better than 2 nm in the band of 420-780 nm, which further proved the effectiveness of the proposed method. In a word, this paper provides a flexible and simple structural design method based on prism-grating combination.

This study designed a ring resonator based on a 2×2 fiber coupler to create a high-performance microwave photonic filter with tunable broadband, single transmission peak, and narrow filter bandwidth. The output characteristics of the resonator were deduced, simulated, and measured. The high-performance microwave photonic filter was realized by combination with an ordinary optical filter. The experimental results show that the filter bandwidth of the designed fiber ring resonator can be as low as 1.2 MHz. It can realize a dynamic selection of any single transmission peak after cascading with the common tunable optical filter. The optical RF signal can be tuned to 40 GHz through a coherent detection link. Compared to that with a traditional noncavity microwave photonic filter, the proposed resonator significantly reduces the filter bandwidth. Meanwhile, compared to the single-fiber ring resonator, the proposed resonator can realize a dynamic selection of any single transmission peak. The designed fiber ring resonator is expected to be used in ultra-fine spectral analysis and optical fiber sensing in the future.

We successfully designed and fabricated fiber Bragg grating (FBG) earth pressure sensors based on additive manufacturing (i.e., 3D printing) technology by embedding an FBG in a 3D printing model made of carbon fibers for pressure monitoring under different conditions. The systematic calibration experiment and stability test of the pressure sensor prepared by additive manufacturing are carried out. The results indicate a good linear relationship between the wavelength and the change in the pressure loaded on the sensors and show that the sensors show good stability in the 100 loading-unloading cycle tests. The experimental results also show that there is a strong correlation between the infilling density of additive manufacturing and key measurement parameters of the pressure sensors, and that the infilling densities of 20%, 40%, 60%, 80%, and 100% correspond to the sensitivities of FBG earth pressure sensors of 0.69, 0.45, 0.39, 0.21, and 0.19 pm/kPa, respectively. In the field tests, all the seven FBG earth pressure sensors (with a depth interval of 2 m and the sensitivities of 15.38, 0.61, 0.15, 0.76, 0.3, 0.15, and 0.13 pm/kPa) can measure the earth pressure, which is consistent with the theoretical value, indicating good performance of the proposed pressure sensors in actual applications. In conclusion, the proposed earth pressure sensors with an FBG earth pressure sensor encapsulated inside the additive manufacturing model, overcome the shortcomings such as the inflexibility of parameter adjustment for traditional sensors, being susceptible to electromagnetic interference, poor environmental corrosion resistance, complex manufacturing process, inevitable assembly errors, and long R&D and production cycle.

To solve the problem of high complexity in the construction of polarization codes in wireless optical communication, a turbulence partial-order construction method with lower coding complexity is proposed, which is suitable for atmospheric weak turbulence channels. In this study, the conventional partial-order technique is introduced into the atmospheric weak turbulence channel, and the relationship among the subchannels of the weak turbulence is determined via Monte-Carlo method simulation. Combining the established relationship with the polarization weight formula, the optimal parameter value is derived. The simulation results show that under different intensities of atmospheric turbulence, the turbulence partial-order method can achieve the same bit error performance as the Monte-Carlo method at low bit rates, and its performance is also similar to that of the Monte-Carlo method at high bit rates, only 0.07-0.2 dB loss is generated at the bit error rate of 10 -4. This is suitable for the improvement of the transmission efficiency of the system and provides a solution for the efficient combination of polarization codes and wireless optical communication.

Segmenting the point cloud from the 3D point cloud data of a subway tunnel is a key step to automatically detect the damage of the subway tunnel and reconstruct a 3D model of the tunnel. The collected 3D point cloud data are inaccurate for calculating the normal vector and curvature of the point cloud because of the structural characteristics of an automated detection system. This renders some common 3D point cloud segmentation algorithms, such as an improved region growing segmentation method, unsuitable for the point cloud data collected by the detection system. To segment the 3D point cloud data collected by an automated detection system, an algorithm based on density clustering was designed and implemented. This algorithm avoids the use of inaccurate normal vector and curvature, overcoming the limitations of an automatic detection system. Finally, we compared the segmentation results of the region growing segmentation method with those of the designed segmentation algorithm based on density clustering using the actual 3D point cloud data.

Aiming at the problems in the semantic segmentation of remote sensing images, such as missed detection of multi-scale targets and rough segmentation boundary, we propose a method of building change detection for aerial images based on an attention pyramid network. The method adopts an encoding-decoding configuration. In the encoding phase, we utilize ResNet101 as the basic network to extract the features and apply dilated convolutions to improve the receptive field in partial residual modules. Meanwhile, the pyramid pooling structure is selected as the last layer of the encoding network to extract multi-scale features of the images. In the decoding phase, the attention mechanism is employed in lateral connection to highlight significant features, and the procedure of top-down dense connection is used to calculate the feature pyramid and then to fuse the features with different resolutions at different phases. Furthermore, the verification experiments are performed on the dataset of building change detection, and the results indicate that our method has good adaptability to different-size-building change detection and has certain advantages in comparison with the classical semantic segmentation networks.

To solve the problems of artifacts and weak edges of industrial computed tomography (CT) images, an image region-scalable fitting energy minimization segmentation method based on wavelet transform is proposed to achieve the accurate positioning of image edges, and improve the image measurement accuracy. First, the wavelet transform is used to preprocess the image in order to reduce metal artifacts. Then, the proposed method is employed to accurately segment the image, which aims to improve the location accuracy of the edge of the region of interest. Actual data measurement results show that the proposed method can effectively reduce the effect on weak edges of the images, and the relative error of measurement is less than 0.7%, which is 1.4 times higher than that of the Chan-Vese algorithm and meets the requirements of measurement applications.

The decrease of the frequency response of the optical transfer function in the optical synthetic aperture imaging system will inevitably lead to image blur. Therefore, it is usually necessary to use the Wiener filtering or blind deconvolution algorithm to achieve image restoration, and clear and high-resolution images are obtained finally. A deep learning frame based on a U-shaped convolutional neural network is proposed. The data set is constructed by the MATLAB software to train the network. The image restoration effects of the trained U-shaped network and blind deconvolution algorithm are compared. The numerical simulation results show that the U-shaped network has strong recovery ability, generalization ability, and versatility in the image restoration based on the optical synthetic aperture imaging system under the condition of weak noise. It can realize fast blind restoration for images and has potential application prospects.

In order to realize the clear imaging of underwater targets with high degree of polarization (HDOP), the traditional polarization imaging model is analyzed. Based on the Schechner polarization imaging model and the definition of polarization degree, a polarization imaging restoration algorithm is proposed to globally estimate the polarization degree of backscattered light, which considers the polarization degree of reflected light of targets. The polarization imaging experiment is conducted, in which the polarization images of three kinds of targets in turbid water with different concentrations are restored. The restoration results show that the measure of enhancement value of the restored image is increased by more than 90% compared to the original Schechner algorithm. Meanwhile, the average gradient, standard deviation and information entropy of image grey are also increased. The restoration results of different polarization images of targets show that the proposed algorithm is not only suitable for the underwater targets with low degree of polarization (LDOP) and rough surfaces, but also makes the underwater targets with HDOP and smooth surfaces gain satisfactory restoration effect.

To quantitatively analyze the influence of image-rotation on the image quality during dynamic drift, a fusion model comprising a static image and dynamic drift is proposed based on the dynamic circular scanning imaging theory. A simulation method of image-rotation vector decomposition and refusion is designed based on the imaging and dynamic motion characteristics of area array CMOS camera. Further, an image-rotation simulation analysis method of the dynamic circular scanning format image is designed. Thus, all simulated images under different dynamic imaging parameters are obtained. Finally, the parameters of image width, ground resolution, signal-to-noise ratio, peak signal-to-noise ratio, and structural similarity are used to quantitatively evaluate and analyze the simulated images. Simulation results show that a longer exposure time of camera corresponds to a higher signal-to-noise ratio of the imaging system, larger image rotation, worse imaging quality, and smaller structural similarity. The results of this research can provide an effective reference for reducing the influence of image-rotation and improving the image quality of the camera in the future design of multiple dynamic imaging modes.

Dual-energy computed tomography (DECT) has been widely used to medical imaging, security inspection, nondestructive testing, materials science and so on, with its capability to decompose and identify materials and provide quantified results. DECT technique can accurately decompose two basis materials due to its performance to acquire the attenuation information of the scanned object at low and high energies. However, when there are three basis materials, if the direct inverse material decomposition (DIMD) is used to decompose the materials, the material CT images will contain much noise and artifacts. Therefore, we propose an image domain multi-material decomposition algorithm for DECT based on dictionary learning (DL) and relative total variation(RTV), which is called DL-RTV for short. The method employs the DIMD to acquire original material images, and then trains a dictionary to explore the sparsity of the images and improve the accuracy of the material decomposition. Meanwhile, the RTV is introduced to further reduce the noise and artifacts of the images and preserve details. In addition, the constrains of mass conservation and the bounds of each pixel are added into the DL-RTV model to enhance the material decomposition accuracy. Simulation and experimental results indicate that the DL-RTV method can decompose three kinds of materials accurately, suppress the noise and artifact of the basis images and improve the material discrimination. The method is authenticated to be effective and practical, which has important significance for the development and application of DECT.

He-Ne gas ring lasers are the core device of laser gyros. In order to avoid the competition effect of the modes operating in the same and reverse directions with rotation in a ring laser, the laser gyro at the single longitudinal mode needs to adjust the composition of Ne isotopes in the gain medium and adopt a dual-isotope gain medium. For this purpose, the Lamb theory and plasma dispersion function were used to study the effects of the frequency splitting of Ne isotopes, isotope composition, inflation pressure, and other factors on the intensity tuning curves of a ring laser. Besides, the experimental platform for the line-type detection of the intensity tuning curves in the laser gyro was set up. At the Ne 20/Ne 22 ratios of 1∶1 and 7∶3 and the inflation pressure of 400 Pa, the intensity tuning curves of the laser gyro were detected, verifying the correctness of the theoretical analysis. In conclusion, this study lays a foundation for the frequency stabilization of laser gyros at the asymmetrical position of frequency with double longitudinal modes.

To solve the fuzzy problem of edge information in mask results by single-stage PolarMask, a contour-point refined network is proposed herein. By predicting the angel offset and distance for each contour point, a more accurate contour can be generated. Moreover, an extra semantic segmentation is added to further refine the edge information. Experiments show that the proposed method achieves a segmentation accuracy of 32.5% on the MS COCO test dataset, 2.1 percentages higher than the fundamental PolarMask, demonstrating the effectiveness of the proposed method.

Owing to imaging principle limitations, depth camera is impossible to obtain the true depth value of a non-Lambertian high reflective surface and transparent substances, resulting in model missing and reconstruction failure. Aiming at the problem of three-dimensional (3D) reconstruction on special surfaces, a method of positioning special surfaces and optimizing position tracking is proposed herein. In the proposed method, positioning for special surfaces is based on the statistics of the zero depth area and timing consistency constraint optimization. The confidence of the depth data is evaluated, and the non-uniform pose calculation method in space and time is used to reduce the influence of transparent materials on pose calculation. The experimental results show that in natural scenes, the proposed method can reconstruct and repair 3D models with special surfaces using only the depth data collected by a consumer camera, which makes pose tracking more accurate and increases model accuracy.

In this paper, a critical pattern selection method for full-chip source mask optimization is proposed. The critical frequency of the pattern represents the characteristics of the pattern. The location and contour information of critical frequency are used to describe the distribution characteristics of the critical frequency in the frequency domain. The corresponding critical frequency extraction method, covering rules, grouping method, and critical pattern selection method are designed to realize the critical pattern selection of the full-chip source mask optimization. Commercial computational lithography software Tachyon of ASML company is used for simulation verification. The comparison with the similar technique of ASML company shows that the process window obtained by this method is better than the ASML Tachyon method.

From the perspective of the actual application of a space camera on the JL-1 satellite, this paper proposed to use high-precision thin-walled tubes and truss rods based on carbon fiber reinforced polymer (CFRP) as the support structure between the primary and secondary mirrors. Besides, related design and engineering analysis are completed. Furthermore, relevant vibration, static and temperature stability tests are carried out in depth. The analysis and test results show that the proposed support structure has good structural stability and resistance to the line of sight jitter and the mass of the carbon fiber truss is only 2.11 kg. Additionally, when the fundamental frequency reaches 186 Hz, the stability under gravity and 4 ℃ temperature rise is better than 1″. The dynamic modulation transfer function (MTF) analysis of the panchromatic remote sensing images obtained in orbit shows that the in-orbit dynamic MTF value of the space camera is 0.08, which further proves the reliability and effectiveness of the truss structure support technology based on CFRP.

The three-layer structure of silver/graphene/silver (AG/GE/AG) was prepared by vacuum evaporation method and wet transfer method, and then through a high temperature annealing, a reusable AG/GE/AG composite structure substrate was obtained. The electric field distribution and theoretical enhancement factor were calculated using COMSOL Multiphysics simulation software. The Raman measurements show that the relative standard deviation (RSD) values of the D, G, and two-dimensional peaks of graphene are 26.0%, 17.8%, and 23.6%, respectively, indicating its good uniformity. The detection limits for rhodamine 6G (R6G) and crystal violet (CV) are as low as 10 -7 mol/L. Using R6G with a concentration of 10 -5 mol/L as the probe molecule and sodium borohydride solution as the cleaning agent, the repeatability experiment of AG/GE/AG substrate was studied. Taking Raman peaks at 613 cm -1and 773 cm -1 as examples, the Raman intensity was maintained at 71.08% and 71.60%, respectively.

A grating-type ultra-wide-band solar absorber was designed based on the refractory titanium nitride (TiN) and titanium dioxide (TiO2). The absorption characteristics of this absorber were studied using the finite element method and the effects of its structural parameters, working wavelength and incident angle on its absorption performance were analyzed. The results show that the absorption characteristics can be effectively controlled by adjusting the structural parameters. In the wavelength range of 500-2000 nm and the incident angle range of 0°-75°, the absorption efficiency is larger than 80%, indicating that the designed absorber possesses the characteristic of ultra-broadband absorption. In addition, the top layer of the unit structure adopts a hemispherical structure, which can increase the average absorption rate of the absorber. The absorber designed in this paper has a potential application prospect in the thermo-photovoltaic and other fields.

The optical surface of sub-wavelength grating structures has anti-reflection characteristics, which are of great significance for the application of photoelectric conversion efficiency. In order to improve the absorptivity of silicon-based grating structures in the near-infrared band (0.78-2.50 μm), we add Ag nanoparticles and an Al2O3 dielectric layer to the gap of the silicon surface grating. Furthermore, the FDTD software is employed to investigate the influence of different composite structures and particle diameters on the light absorptivity and analyze the light intensity distribution of the characteristic section in the composite structure at different working wavelengths. The experimental results show that in the context of a line/space ratio of 1∶1 and a period of 1 μm for the grating, as well as two-layer Ag particles (diameter of 0.25 μm) cover with an Al2O3 dielectric layer in the periodic groove, the average absorptivity of the composite microstructure in the near-infrared band can theoretically reach 0.463, achieving absorption enhancement. It follows that the metal-silicon gratingdielectric layer composite structure can enhance light absorption, in the application of photoelectric conversion devices, thereby improving photoelectric conversion efficiency.

GF-6 WFV is a high-spatial resolution multi-spectral sensor loaded on Chinese GF-6 satellite, which realizes the combination of high spatial resolution and wide coverage. Accurately identifying the cloud pixels of GF-6 WFV data is of great significance for supporting agricultural resources monitoring, forestry resources investigation, disaster prevention and mitigation and other industry applications. Based on the global land cover product—FROM-GLC10 (Finer Resolution Observation and Monitoring-Global Land Cover 10) data, the LCCD (Land Cover-based Cloud Detection) algorithm is improved to carry out cloud detection of GF-6 WFV data in the paper. Taking FROM-GLC10 data as a priori data, fully considering the change of reflectivity of different surface types, different methods are used to set thresholds for each surface type. The accuracy of cloud detection results was evaluated by visual interpretation, and the cloud accuracy rate as a whole reached 92.46%, among which the cloud accuracy rates of vegetation type, water type and highlighted surface type were 93.09%, 95.60% and 88.70%, respectively. The results show that the improved cloud detection algorithm based on surface type effectively improves the accuracy of cloud detection of GF-6 WFV data.

The attention mechanism of neural networks can extract key information from data, and the application of this feature in the selection of hyperspectral bands can help fully learn the interdependence and nonlinear relations between bands and extract more important bands. This paper presents a multi-objective optimization method for hyperspectral band selection based on the attention mechanism. First, the attention module and autoencoder are used to construct the network. Then, one-dimensional spectral data is provided as input to the network; two loss functions are used and combined with the multi-objective optimization method for training. Therefore, the attention module embedded in the network learns the nonlinear relationship between different bands and assigns more weight to the bands with a large amount of information and easy classification, thereby realizing band selection. Finally, the support vector machine classifier and mean spectral divergence are used to validate the performance of the band subset. The experimental results show that the band subset extracted using this method from the Botswana and Indian Pines datasets is more accurate and informative than the subsets extracted using other algorithms. Thus, it is demonstrated that this algorithm is more effective in selecting hyperspectral bands.

In this paper, the infrared radiation characteristics of the gas-solid two-phase exhaust plume containing solid particles are simulated and modeled. The ray tracing method is used to calculate the radiation intensity and radiation suppression rate of the exhaust plume. The simulation environment of the exhaust plume flow field and the imaging simulation scene of the air-based infrared camera are designed. Through the mechanism of solid particle radiation suppression, the particle diameter, particle flow rate, and particle complex refractive index are determined as the main influencing factors of the radiation suppression rate. Simulation calculation results show that reducing the particle diameter and increasing the particle flow rate can enhance the radiation suppression ability of the solid particles. Increasing the particle absorption cross-section can enhance the radiation suppression ability of solid particles in the gas absorption waveband, and at the same time reduce their radiation suppression ability in the non-gas absorption waveband. In some cases, the solid particles may increase the radiation intensity of the exhaust plume and change the spectral radiation characteristics.

This paper developes a set of automatic platform for testing scattering characteristics and uses this test platform for the experimental measurements on the surface scattering characteristics of stray light absorbing materials in a Thomson scattering diagnostic system. In the 1064 nm band, the noise floor of the test platform for the bidirectional reflection distribution function is better than 2×10 -5 sr -1. Furthermore, this test platform is employed to measure the surface scattering characteristics of standard white plates, stainless steel, black aluminum alloys, black paints, and other materials. The experimental results show that the Thomson scattering diagnostic system on the “Keda Torus” setup adopted Avian Black-S black paint and Metal VelvetTM black foil to absorb stray light. In the end, the measurement results are utilized to provide ABg model parameters for a variety of materials, offering necessary data for the ray tracing simulation associated with stray light.

To investigate water vapor, methane (CH4), and carbon dioxide (CO2) concentrations in the atmosphere, two near-infrared narrow band-pass filters are fabricated on sapphire substrates. The central wavelengths are 1375 nm and 1610 nm and bandwidths are 15 nm and 60 nm, with average transmission of 95%. Nb2O5/SiO2 Fabry-Perot multilayer filters are optimized and deposited using dual ion beam sputtering. Compared to electron beam evaporated films, the dual ion beam sputtered films have significantly improved surface quality and reduced defects, and their root-mean-square surface roughness is reduced to below 1 nm, which significantly improves the uniformity of electric signals when coupled with a photodetector.

Based on the chalcogenide glass IRG206 (As40Se60) substrates, we prepare the 8-12 μm anti-reflection protective films. This kind of film contains two parts: dielectric film and diamond-like carbon (DLC) protective film. It can be applied in the window lens of the non-thermalized infrared detection system. Moreover, the fracture mechanism of dielectric films is studied, and the deposition temperature of dielectric films is optimized to improve the adhesion of dielectric films. In addition, the effect of pressure on the microstructures and stress of DLC films is studied. By splitting the DLC film into the low-stress adhesive layer and the high-stress friction-resistant layer, the film stress is reduced, but the friction resistance as well as firmness of films are improved. The test results reveal that the average reflectivity of anti-reflection films at 8-12 μm is 3%. Furthermore, the anti-reflection films pass through the salt spray test, high and low temperature test, and severe abrasion test specified by GJB2485-1995, and possess good environmental adaptability.