With the rapid development of offshore oil exploration and development technology, an increasing amount of production water needs to be treated by water treatment facilities before being discharged into the sea or reinjected, posing a serious threat to marine environmental safety. Establishing a comprehensive optical in-situ monitoring technology for pollutants discharged from offshore oil platforms is of great significance for promoting the development of smart oil fields and ensuring marine ecological security, in order to facilitate efficient exploration and development of marine oil and gas. Although multiple core technologies have been breakthrough in the field of marine environmental monitoring sensors, online monitoring of oil in water still mainly remains in the stage of laboratory prototype verification and theoretical exploration, and has not yet fully adapted to the special requirements of harsh working conditions on offshore platforms. Two types of oil sensors in water were designed based on ultraviolet fluorescence method, with measurement accuracy better than 0.1 mg/L in the range of 0~100 mg/L. The linear fitting coefficient of the sensors reached R2>0.97.The relative stability tests for different concentrations all meet RSD<0.5%, which meets the requirements for offshore field applications. Afterwards, an online monitoring system for oil content in the production water of offshore oil platforms was built and sea trials were completed. During the experiment, the online monitoring system for oil in water achieved 24-hour continuous acquisition of oil content data in the production water discharged from offshore platforms. The experimental results show that the developed fluorescence based oil monitoring system in water can accurately reflect changes in oil concentration in water, and has fast response capability and high concentration detection range for substances in water. Considering that the transportation of platform water samples back to the onshore laboratory for testing does not have real-time capability, and the water samples may undergo physical and chemical phenomena such as sedimentation and dissolution, and can not reflect changes in oil concentration in the production water over a period of time, a one month online monitoring and laboratory manual testing method comparison test was conducted at the offshore site. The results indicate that the two methods have a high degree of trend consistency, with test data reaching 98.5% within the standard range. This suggests that using fluorescence method for on-site online monitoring has the characteristics of simple operation and high accuracy, meeting the requirements of offshore field applications. This method will greatly improve the efficiency and data accuracy of water monitoring for offshore platforms, and also provide reliable data support for the unmanned operation of offshore platforms. This research and development plan is selected as a demonstration application area for a certain platform in the East China Sea, which has the same applicability for different sea areas and oil fields. In the later stage, a large sample range will be used to conduct relevant experiments on offshore platforms in multiple regions to verify the feasibility of fluorescence based online detection of oil concentration in water. This technology will also serve as a key support for the construction of smart oil fields, ensuring real-time protection of marine ecological security.

In the real world, weather conditions rarely present themselves in a simple, singular, and uniform manner. Instead, they usually exist in a mixed state of countless randomly distributed particles. These particles are of a wide variety. In natural environments, we encounter elements such as rain, fog, or snow. In the realm of man-made environments, there are randomly dispersed particles like smoke. Regardless of their origin, these particles have the potential to exert a significant and far-reaching impact on laser transmission.When a laser signal propagates through this complex particle environment, it undergoes a substantial attenuation process. As it passes through numerous discrete and randomly arranged particles, its energy is continuously absorbed and scattered. This complex process transforms the laser echo signal into an extremely complex and unstable one. The consequences are two-old: not only does it exponentially increase the complexity of signal processing, but it also significantly weakens the reliability and stability of the laser communication system. Therefore, dealing with the impact of these complex particles on laser transmission has become a core and crucial challenge in multiple key fields, including the fields of laser communication, target recognition, navigation, and fuze technology.This study comprehensively explores the combined effects of a variety of particles (rain, fog, snow, ammonium sulfate, silicate, and black carbon) across a range of wavelengths. It carefully takes into account the rich and diverse meteorological conditions prevalent in the real-world environment. By establishing laser transmission models for three different mixed particle groups (each particle group having unique distributions and complex refractive index characteristics), this study has gained a deeper and more profound understanding of these complex phenomena. Utilizing the powerful tools of second-and fourth-order speckle statistics, an in-depth and detailed analysis has been carried out on the intensity fluctuation correlation function and the intensity distribution autocorrelation function of speckle patterns at different transmission distances.The results of this study have led to a series of key and enlightening observations. First of all, for the three mixed particle swarm models under study, as the transmission distance gradually increases, the scattering intensity undulation correlation function shows a continuous upward trend. Notably, when considering the 1 064-nanometer wavelength, under the influence of larger particle swarms, this function only experiences relatively small and subtle changes. It turns out that different particles have different and unique effects on the scattering intensity undulation correlation function. In scenarios consisting solely of rain, fog, and snow particles, raindrop particles become the dominant factor influencing the overall variation of this function. Conversely, in environments filled with ammonium sulfate, silicate, and black carbon particles, for 532-nanometer laser light, the correlation between different positions within the same spatial region is significantly weakened. This weakening of the correlation can be attributed to the stronger penetration ability of near-infrared wavelengths, which in turn leads to distinctly different scattering behaviors.The sharpness of the peaks in the autocorrelation function of the scattering spot's intensity distribution is also intricately related to the proportion of different particles. Specifically, the larger radius of raindrop particles results in more widespread and dispersed light scattering, while the significant imaginary part in the complex refractive index of black carbon particles leads to more intense absorption of laser light. Among all the particles analyzed, an increase in the proportion of ammonium sulfate particles has the most significant and far-reaching impact on this function, followed closely by raindrop particles. By carefully adjusting the contrast of the simulated scattering patterns, these patterns can be clearly distinguished visually. A comparison of these patterns vividly shows that the scattering pattern generated by the 1 064-nanometer laser wavelength becomes more obvious and prominent after passing through a random medium filled with ammonium sulfate particles.In such a complex and unpredictable random environment, the information transmitted through the LiDAR system is extremely vulnerable to interference. The insights obtained from this study provide valuable and crucial guidance for optimizing LiDAR signal coding and modulation. By thoroughly understanding how various environmental factors and particle characteristics affect laser transmission, it is possible to enhance the anti-interference ability of the transmitted information in these challenging environments, thereby ensuring its accuracy and stability. These insights also provide the necessary and crucial theoretical support for extracting accurate and effective target information. In addition to optimizing communication systems, a deep understanding of the relationship between the scattering intensity distribution and particle swarm characteristics can completely revolutionize the optimization process of feature extraction algorithms for point cloud data. Therefore, this study has laid a solid and stable foundation for subsequent important tasks, such as creating accurate three-dimensional models and identifying specific static or dynamic targets. Ultimately, these research results play an important role in improving the efficiency and reliability of LiDAR systems and related technologies in practical applications, thus promoting innovation and improvement in a wider range of fields.

Light-homogenizing devices, as key components in optical systems, directly influence the overall performance of these systems. With the continuous advancement of science and technology, research and development of light-homogenizing devices have garnered increasing attention. Among various light-homogenizing technologies, Holographic-polymer Dispersed Liquid Crystal (H-PDLC) volume gratings have shown significant potential in display technology and lighting systems due to their unique advantages, such as thinness, low-cost fabrication, and light tunability, offering superior solutions for the design of light-homogenizing display devices. These characteristics of H-PDLC gratings make them highly applicable in fields like display technology and lighting systems. However, to achieve H-PDLC gratings with high diffraction efficiency and transmittance, in-depth research is required on the selection of holographic materials, the phase separation of polymeric liquid crystals, and the grating morphology. Researchers worldwide have made certain achievements in these areas, but the application and development of two-dimensional H-PDLC gratings still face challenges, particularly in terms of diffraction efficiency and structural stability. To address these issues, this study derived a three-wave interference holographic theory and designed a corresponding three-wave interference holographic optical path. Utilizing COMSOL Multiphysics software, a refractive index model corresponding to a one-dimensional grating formed by three-wave interference was constructed, thereby validating the feasibility of using the three-wave interference holographic optical path to prepare one-dimensional gratings. In the experiments, by precisely controlling the angle of the reflecting mirrors in the exposure optical path, the exposure angle between the beams was managed, leading to the successful fabrication of H-PDLC one-dimensional gratings with three distinct grating periods, corresponding to diffraction angles of 10.4°, 14° and 18.4°, with an angle error kept within ±0.5°. In terms of fabrication technology, this study successfully prepared H-PDLC gratings on Tri-cellulose Acetate (TAC) membranes and created a two-dimensional grating film with two mutually perpendicular one-dimensional gratings through a superposition method. The advantage of this method is that the two independently formed grating morphologies are more stable during the interference process, effectively avoiding diffusion-related defects that might occur during the fabrication process. To produce gratings over a larger area, the study further explored the relationship between exposure power and grating diffraction efficiency. Results indicated that as the exposure power decreased, so did the diffraction efficiency of the grating. This phenomenon is attributed to the reduced energy absorbed by the photoinitiator, leading to slower monomer polymerization rate and liquid crystal diffusion rate, which in turn fails to form a well-defined phase-separated structure in the holographic grating. Based on these findings, the study optimized monomer materials by comparing the one-dimensional grating diffraction efficiency under low-power exposure conditions for two types of acrylate monomer systems: PDDA (polymerizable oligomer with acrylate groups) and TMPTA (trimethylolpropane triacrylate). The results showe that under the same exposure conditions, the TMPTA system achieves a first-order diffraction efficiency of 40.43%, while the PDDA system reaches 63.33%, significantly outperforming the TMPTA system. This is believed to be due to the higher molecular weight of PDDA, which allows for the formation of a more compact and stable polymer chain structure during polymerization, resulting in a higher quality polymer network that enhances the diffraction efficiency of the grating. To further enhance the grating's diffraction efficiency, the study optimized the ratio of monomer PDDA to film-forming resin EP828. SEM inspections were conducted to observe the surface morphology of the gratings under different material ratios. The results indicate that increasing the PDDA content appropriately improves the phase separation within the grating structure, thereby enhancing the diffraction efficiency. When the ratio of PDDA to EP828 was adjusted to 2∶1, with an exposure power of 5 mW/cm2, the H-PDLC grating achieved a first-order diffraction efficiency of over 90%. The final two-dimensional grating film had dimensions of 50 mm×50 mm, with a transmittance exceeding 90% across the visible light spectrum. This achievement not only provides a new theoretical approach to the design of H-PDLC gratings but also experimentally verifies their high performance, laying a solid foundation for further research and application of H-PDLC gratings. In summary, this research successfully addresses the challenges related to diffraction efficiency and structural stability of H-PDLC gratings through rigorous theoretical analysis and experimental validation. By optimizing the material composition and exposure parameters, this study significantly improves the diffraction efficiency and light homogeneity of H-PDLC gratings, which is of great significance for promoting their application in display technology and lighting systems. Future research can further explore various material combinations and fabrication processes to achieve H-PDLC gratings with even higher performance, meeting the growing demands of optical systems.

Extrinsic Fabry-Pérot interferometer sensors are widely applied in aerospace, biomedical, electric power, and hydroacoustic detection fields due to their high sensitivity, rapid response, electromagnetic interference resistance, compact size, and low cost. Common demodulation methods for fiber optic EFPI sensors in engineering applications include intensity demodulation, spectral demodulation, correlation, phase-generated carrier, and three-wavelength orthogonal phase demodulation. The three-wavelength demodulation method, characterized by fast demodulation speed, large range, strong anti-interference capability, simple principle, and easy miniaturization, has gained increasing attention and research in recent years. This technology uses three different wavelengths of interference light signals for phase extraction in EFPI sensors. However, imbalances in the three light signals due to various inconsistencies during signal transmission, such as fiber bending loss, optical filter loss, light source flatness, sensor spectral envelope flatness, and photoelectric conversion gain, can lead to signal distortion and significant errors in demodulation results. Amplitude normalization can be applied for calibration of three-wavelength interference signals with phase changes exceeding 2π, but calibration is challenging when sensor phase changes are less than 2π. Particularly in high-frequency signal testing, the sensor's resonance frequency must be much higher than the measured signal, leading to lower sensitivity and phase changes far less than one phase cycle. For small signals less than 0.2π, the decrease in signal-to-noise ratio makes calibration even more difficult. Building on our research group's study of three-wavelength dynamic demodulation algorithms, this paper proposes a three-wavelength small-signal calibration method based on Lissajous curve fitting. By systematically enumerating calibration coefficients and adjusting the light intensity of three-wavelength interference signals, the method evaluates the orthogonal fitting assessment of the Lissajous curves for the two arctangent intermediate variables in the three-wavelength demodulation algorithm. It judges the contour deviation and data point density under different calibration coefficients, ensuring the curve fits a circle centered at the origin for accurate small-signal calibration. A demodulation system was established based on the dynamic signal calibration and demodulation needs of fiber optic EFPI sensors. Three equal-power distributed feedback lasers with central wavelengths of 1 553.33 nm, 1 557.36 nm, and 1 561.42 nm were used as the light sources. The three different wavelengths of light signals completed interference modulation in the fiber optic EFPI sensor through a three-channel dense wavelength division multiplexing optical filter and coupler. The reflected light signals, after filtering through the optical filters, were collected by the three-wavelength demodulation board card using photodiodes and analog-to-digital converter modules, converting the three different wavelengths of interference light signals into digital signals for computation in the upper computer system. Simulation results show that, based on the conditions of the three-wavelength demodulation algorithm, this calibration method can stably obtain calibration coefficients for phase changes within 0.2π for small signals, with a maximum amplitude error not exceeding 1.1% and strong noise resistance. Experimental tests indicate that the maximum error in sensor sensitivity after calibration does not exceed 0.15%, and the maximum nonlinear error of the fitted curve does not exceed 0.6%. This calibration method achieves accurate calibration for small signals in the three-wavelength phase demodulation algorithm, enhancing the range and precision of the three-wavelength algorithm.

In recent years, with the growth of video and multimedia applications, the upgrading of radar and electronic warfare systems, and the expansion of satellite communication and navigation systems, the requirements for the bandwidth, real-time spectrum analysis, acquisition, and reliability of Radio Frequency (RF) signals have been continuously increasing. As a key basic technology, broadband RF signal processing is widely applied in data transmission, electronic countermeasures, cognitive radio, and other fields. Traditional electrical methods process RF signals in the electrical domain, meeting conventional signal processing requirements to a certain extent. However, these methods are restricted by inherent electronic bottlenecks such as bandwidth, electromagnetic interference, power consumption, and reliability. Microwave photonics, benefiting from characteristics such as large bandwidth, electromagnetic interference resistance, low transmission loss, and high frequency bands, provides a new solution for broadband RF signal processing.The current main technical schemes can be divided into optical heterodyne detection, photonic-assisted channelization, and microwave photonic filters. The optical heterodyne detection method converts RF signals into beat frequency signals by mixing signal light and local oscillator light. Based on a sampling frequency of 200 kHz, the theoretical bandwidth of the system is 100 kHz, and the resolution can reach 390 Hz. However, frequency drift or phase fluctuation will lead to instability of the beat frequency signal, affecting the measurement accuracy. The photonic-assisted channelization scheme divides broadband signals into multiple sub-channels with small bandwidths. The photonic-assisted channelization scheme using dispersion delay can process microwave signals with an instantaneous bandwidth of 4.86 GHz and achieve a spectral resolution of 340 MHz, but it has high requirements for the performance of dispersive media. The microwave photonic filter method performs filtering processing on RF signals by designing microwave photonic filters with specific frequency responses, which can realize spectral analysis with a spectral range of 0~20 GHz and a resolution of 650 kHz. However, the filter fabrication process is relatively complex and requires high processing precision. In the current research field, how to achieve an optimal balance between bandwidth and spectral resolution has gradually become a key focus and frontier hot issue in related research directions such as broadband RF signal processing. Many research teams have carried out studies from different perspectives to break through existing technical bottlenecks, achieve collaborative improvement of bandwidth and spectral resolution, and further meet the growing demand for processing capabilities of large-bandwidth signals and high spectral resolution. Among them, the technology of processing RF signals based on the spectral hole burning effect has shown significant advantages.Based on the spectral hole burning effect of Tm3+:YAG crystal for broadband radio frequency signal processing, this paper studies the influence of power spectrum writing and reading parameters on spectral resolution. By numerically solving the two-level optical Bloch equation using the Runge-Kutta method, the population distribution changes of each energy level are calculated. Under the condition of inputting a linearly chirped pulse, the crystal absorption spectrum is obtained by using the Beer-Lambert type Fourier domain transfer function method. The signal power spectrum is obtained through differential detection. The influences of crystal homogeneous broadening, writing signal intensity and reading signal chirp rate on spectral resolution are analyzed and discussed. The results show that the spectral resolution increases with the increase of crystal homogeneous broadening, writing signal intensity and reading signal chirp rate. The optimal optical field chirp rate for reading spectral hole burning is 1 MHz/μs. When the crystal temperature is 9 K, the experimentally measured spectral resolution is 300 MHz.

With the continuous development of computer vision research, the application of deep learning has become more and more widespread. The introduction of deep learning into object detection research has greatly improved detection performance. However, in order to further improve the detection accuracy of the model, the depth and width of the network are continuously increased, which leads to models containing a large number of parameters and complex structures, posing challenges for practical deployment. Aiming at the problems of high computational cost, high memory consumption, and difficulty in efficient deployment on edge devices of visible-infrared dual-modal fusion detection models, this paper proposes a lightweight pedestrian-vehicle detection model based on visible and infrared modality fusion. The multimodal input greatly improves the stability of the algorithm during all-weather operation, and the detection task can be well accomplished regardless of snowy, foggy, or low-illumination conditions. The proposed model uses the MobileNetV2 lightweight network instead of the YOLOv7-tiny network backbone. The MobileNetV2 network employs an inverted residual structure and a linear bottleneck layer, both of which effectively enhance its feature representation and learning capability. It also uses depthwise separable convolution, which, unlike conventional convolution, divides the convolution into pointwise and depthwise components. This paper also proposes a differential modal fusion module inspired by the principle of differential amplification circuits. This module differentiates the two modal images, extracts the differential and common mode information, and amplifies the differential mode information to fully utilize the complementary advantages of visible and infrared modalities. The illumination-aware module is introduced due to the fact that under low illumination and bad weather conditions, infrared and visible images have different impacts on model performance. This module dynamically assigns weights to visible and infrared features based on illumination conditions, thus maximizing the feature information of each modality. Three public datasets are used for the experiments: FLIR ADAS, LLVIP, and KAIST. We conduct comparative experiments between lightweight single-modal algorithms such as YOLOv5s and YOLOv7-tiny, and dual-modal algorithms such as ICAFusion and CFT. The results show that the bimodal detection models outperform the unimodal detection models in terms of detection performance. On the FLIR ADAS dataset, the proposed model improves the detection accuracy by 11.6% and 15.3%, respectively, compared to the unimodal YOLOv5s and YOLOv7-tiny models with RGB input. Compared to the unimodal YOLOv5s and YOLOv7-tiny models with infrared input, the detection accuracy is improved by 3.3% and 4.9%, respectively. Compared with the baseline model, the accuracy of the proposed model is improved by 3.8%. Compared with the bimodal models ICAFusion and CFT, the proposed model improves the detection accuracy by 3.8% and 1.9%, respectively. On the LLVIP dataset, the proposed model improves the detection accuracy by 6.9% and 1.1%, respectively, compared to YOLOv7-tiny when using visible and infrared unimodal inputs. The proposed model improves the detection accuracy by 5.4% and 2.0%, respectively, compared to YOLOv5s with visible and infrared input. Compared with the baseline model, the accuracy is improved by 1.3%. Compared with the bimodal models ICAFusion and SLBAF, the proposed model improves the detection accuracy by 9.6% and 1.5%, respectively. On the KAIST dataset, the proposed model improves detection accuracy by 26.9% and 7.8% on visible and infrared inputs, respectively, compared to YOLOv7-tiny. Compared with the baseline model, the accuracy is improved by 3.4%. Compared to other bimodal models, the proposed model shows the most advantageous results in terms of detection accuracy, achieving as high as 76.2%. In terms of inference speed, the proposed model achieves 208 FPS, 103 FPS, and 113 FPS on the FLIR ADAS, LLVIP, and KAIST datasets, respectively. These results show that the proposed model has significant advantages in both detection accuracy and speed, as well as robustness.

Multi-primary-color display technology has been an important technical route to achieve color gamut expansion, which improves display brightness and energy efficiency by adding white sub-pixels. To drive Red-Green-Blue-White (RGBW) displays, color gamut mapping is required. In order to make the color saturation theoretically consistent before and after color gamut mapping, the white sub-pixel grayscale value is usually deducted from values of chromatic sub-pixels in traditional algorithms, which greatly reduces the contribution of chromatic sub-pixels to brightness enhancement, resulting in a less overall brightness increase. In addition, there is currently few objective and intuitive evaluation index to evaluate the performance of algorithms.A mathematical model of color gamut mapping transformation was established by combining the general method of color gamut mapping in this paper. Two key parameters of the mapping rate-of-change function and the white sub-pixel driving function are proposed. According to the characteristics of RGBW color gamut mapping algorithm, a four-in-one evaluation index system was proposed, it includes the Average Brightness Magnification (ABM), the Average Chromatic Sub-Pixels Magnification (ACSM), the Average White Sub-pixel Ratio (AWSR), and the Average Saturation Ratio (ASR). Based on the evaluation index system, the Mapping Rate-of-Change Function and the White Sub-Pixel Driving Function were determined. A linear algorithm function was constructed based on evaluation index system, which was continuously adjusted to a quadratic function to reach the best effect, and a more efficient and simplified new color gamut mapping algorithm was realized. The proposed algorithm is designed to increase the contribution of chromatic sub-pixels to the brightness improvement of the image without deducting the white component and with maximizing the chromatic sub-pixels grayscale values, and it can also balance the AWSR and the ASR by adjusting the weight factor.Simulation images processed by the proposed algorithm is displayed on an RGB Flat Panel Display (FPD) to simulate algorithms effect. Furthermore, the physical verification is carried out on a projector that supports RGBW projection, and the imaging brightness under two platforms is measured respectively by using an image luminance colorimeter. The results show that the brightness of the proposed algorithm can be increased by 2.26 times and 2.59 times on flat panel display and projector platforms respectively. In terms of evaluation indices, the ABM of the proposed algorithm reached 2.304, the ACSM reached 1.677, the ABM reached 2.3 times that of traditional algorithms, and the contribution of the chromatic sub-pixels grayscale values to the brightness improvement reached 1.67 times that of traditional algorithms. The AWSR was 0.261, and the ASR was 0.733, which better balanced the overall brightness enhancement and saturation decrease of images. Simulation rendering pictures of each algorithm on an RGB flat panel display were also taken in the experiment, as well as actual pictures projected by an RGBW projector. Judging from these pictures, the proposed algorithm can significantly enhance the display brightness and have a good performance of color retention. The calculation results of evaluation indices, measurement results and simulation verification results are basically consistent, so the evaluation index system can better show the performance of the RGBW color gamut mapping algorithm. This study provides a reference for evaluation performance of RGBW algorithms and construction of new algorithms in the future, and providing new directions for color gamut mapping in multi-primary-color displays.

In active compensation near-infrared microscopic interferometry for measuring silicon-based High Aspect Ratio (HAR) structures, when the focal plane of the objective lens does not coincide with the trench bottom, applying under-compensated or over-compensated defocus to the deformable mirror causes shifts in the interference signal envelope, thereby affecting the measurement accuracy of depth and other critical topographical parameters. To address this challenge, this study aims to quantify the impact of under-compensated and over-compensated defocus errors on the depth measurement accuracy of HAR silicon trench structures and establish tolerance criteria for defocus aberration compensation. To achieve the above objectives, this study constructs a Three-Dimensional Point Spread Function (3D-PSF) model by combining the Finite-Difference Time-Domain (FDTD) method with angular spectrum diffraction theory. The modeling approach divides the entire optical field modulation process into three stages: First, the FDTD method is utilized to accurately calculate the propagation of the optical field within and on the surface of the structure, thereby obtaining the near-field optical field distribution modulated by the sample structure surface. Second, the angular spectrum diffraction theory is employed to propagate the near-field optical field to the image plane of the microscopic objective lens, aiming to describe the optical field transmission characteristics during the microscopic imaging process. Finally, the test optical field and the reference optical field are extracted from the optical field on the image plane. Interference information is obtained through coherent superposition, and the 3D-PSF is generated by stacking axial scans. Based on the generated 3D-PSF, an inverse filter is further designed to correct the phase distortion introduced by high aspect ratio structures in the frequency domain, simulating the aberration compensation process of the deformable mirror. On this basis, the study conducted simulations on three structures with gradually increasing aspect ratios (depth of 50 μm and line width of 10 μm; depth of 100 μm and line width of 10 μm; depth of 200 μm and line width of 15 μm) under different defocus conditions. The interference signal envelopes at the bottom of the trenches were analyzed, and the depth measurement errors were evaluated. The results showed that the variation trends of the interference signal envelopes of samples with different aspect ratios were consistent when the defocus amount changed in the same way. This indicates that the impact of under-compensated and over-compensated defocus errors on depth measurement has a weak correlation with the geometric properties (such as depth and line width) of the samples. In addition, within the range where the defocus error is less than ±3λ, the offset of the envelope peak is consistent with the variation of the depth measurement results. In this case, the error is small, and the relative error of depth measurement is less than 1%. However, when the defocus error exceeds ±3λ, the peak broadening phenomenon significantly increases the error of the coherent peak positioning algorithm based on the centroid method. To verify whether the influence law of defocus errors on depth measurement results based on the 3D-PSF inverse filter obtained from simulation analysis and the tolerance criteria for defocus errors are consistent with the experimental results, defocus error compensation experiments were conducted on a silicon-based trench structure with a nominal depth of 101.77 μm and a nominal line width of 10.97 μm using the self-developed active compensation near-infrared microscopic interferometry system. The experimental results are consistent with the simulation results. When the defocus error is within the range of ±3λ, the relative error of depth measurement is less than 1%, and its impact on the measurement results is negligible. This study provides comprehensive simulation and experimental data support for the quantitative analysis of the influence laws of defocus errors by combining simulation and experimental methods. The proposed tolerance range of defocus errors offers specific theoretical reference for the optimized design and precision control of microscopic interferometry systems.

From integrated optical circuits to long-haul fiber-optic networks, most photonic devices and communication systems rely on low-loss optical fibers. One of the key parameters affecting the transmission performance of optical fibers is their refractive index distribution, which determines characteristics such as insertion loss, propagation modes, and bandwidth of optical fibers. However, internal defects within the fiber can cause scattering or absorption of the optical signal during transmission, leading to attenuation and leakage of the output optical signal. Therefore, accurately and rapidly measuring the refractive index distribution and internal defects of optical fibers is of great significance for the optimization of fiber structure design and quality monitoring. Currently, the main methods for measuring the refractive index of optical fibers include the refractive near-field method, atomic force etching method, focusing method, thin-film interferometry, and transverse interferometry. Among these, transverse interferometry based on microscopic imaging requires no pre-treatment of the fiber. The fiber can be directly immersed in a matching liquid to achieve rapid, non-destructive measurement of the three-dimensional refractive index of the fiber. Therefore, this paper aims to construct a transverse interferometric system to reconstruct the three-dimensional refractive index distribution of optical fibers, enabling the detection of internal geometric structures and defects within the fiber. Based on the method of microscopic interferometry, this paper designs a cylindrical lens system for fiber measurement that effectively compensates for imaging astigmatism and proposes a high-precision transmissive transverse microscopic interferometric tomography system and method based on the cylindrical lens. First, a fiber simulation model was established based on the Finite Difference Time Domain (FDTD) theory for simulation verification. The results showed that the refractive index reconstruction error was extremely small, verifying the feasibility of the measurement method. Secondly, regarding the selection of the refractive index matching liquid in the experiment, the refractive index distribution of the fiber was reconstructed through single-direction projection to determine the optimal refractive index difference between the matching liquid and the fiber cladding. On this basis, three-dimensional refractive index tomographic reconstruction experiments were conducted on both multimode and single-mode fibers, and the results were compared with those obtained using a spherical lens. The results showed that the measurement error of the fiber core diameter using the cylindrical lens system was reduced by a factor of 10, and all measured values were within the given error range. Furthermore, the fiber defect detection experiment results indicated that the size, shape, and location of internal fiber defects could be identified through the reconstruction of the three-dimensional refractive index distribution of the fiber.

In recent years, composite materials have made remarkable progress in improving production efficiency, reducing manufacturing costs, and optimizing overall performance, and they have been successfully applied in engineering practice. However, during manufacturing and service, composites are prone to internal defects such as delamination, debonding, and cracking, which can lead to structural failure. These defects degrade material performance and pose serious safety concerns. Therefore, developing efficient and reliable Non-Destructive Testing (NDT) techniques is crucial to ensure the quality of composite materials. As an important optical NDT method, shearography can accurately capture the shape and location of defects. With advantages such as non-contact measurement, high sensitivity, and full-field detection, shearography shows great potential for detecting defects in composites.Due to significant differences in the physical properties of different composite materials—such as thermal conductivity and stiffness—their responses to external loads also vary considerably. As a result, the choice of loading strategy has a significant impact on detection accuracy and the recognition accuracy of defect fringe patterns. However, current loading strategies largely depend on the operator’s experience, lacking systematic parameter selection guidelines. This often leads to overloading or underloading, thus affecting the detection performance.To evaluate the influence of different loading strategies on defect detection effectiveness, this study selects two typical loading methods—thermal loading and vacuum loading—to perform quantitative loading experiments on various composite materials. Phase maps of defective regions are obtained using shearography. After filtering the results from both loading methods, a YOLOv9-based defect phase map recognition model is constructed to intelligently identify and analyze the detection images. By comparing the recognition rates of defect fringe maps under different loading strategies, the most suitable loading strategy for each material is determined.The experimental results are as follows for aluminum honeycomb panels, the highest defect detection rate under thermal loading is 38.26%, while under vacuum loading it reaches 91.3%; for carbon fiber-reinforced composites, the highest detection rate under thermal loading is 97.28%, while only 18.9% under vacuum loading; for aluminum skin-paper honeycomb panels, the detection rates are 38.26% for thermal loading and 91.27% for vacuum loading.These results demonstrate that for metal-based composites, due to their high thermal conductivity, thermal loading cannot effectively stimulate deformation differences between defective and intact areas, leading to poor detection results. In contrast, vacuum loading produces significantly better outcomes. On the other hand, for carbon fiber-reinforced composites, which possess low thermal conductivity and high stiffness, the temperature gradient between defective and intact regions is relatively large, enabling thermal loading to more effectively highlight defect features. In terms of loading magnitude, the experiments show that at the initial stages of loading, the amount of deformation is small, resulting in faint fringe patterns and low defect recognition rates. As the loading level increases, the fringe contrast becomes clearer, and recognition performance improves. However, when the loading becomes too high, the fringe density becomes excessive, which adversely affects recognition accuracy. Notably, the peak recognition rate under different loading levels varies across material types. For composites with lower stiffness, excessive loading can lead to overly dense fringes, reducing detection effectiveness. Therefore, it is essential to select appropriate loading magnitudes based on the stiffness characteristics of each material.The findings of this study provide both theoretical guidance and practical reference for selecting loading strategies in shearography-based NDT, and are of significant value in improving the accuracy and automation level of defect detection in composite materials.

To address the application requirements for relative radiometric calibration and image non-uniformity correction of multi-CCD cameras during on-orbit operations, this study proposes an LED point source-based relative radiometric calibration method for spaceborne systems. Grounded in the inverse-square law of point source irradiance, this methodology employs spatially fixed stable light sources as radiometric references. By establishing a comprehensive technical framework encompassing laboratory cross-comparison calibration, on-orbit calibration coefficient correction, and performance verification, it achieves full lifecycle monitoring and correction of relative radiometric calibration for remote sensing payloads.Through laboratory analysis of the spectral irradiance spatial distribution characteristics of LED point source modules, we established the correlation between light intensity spatial distribution and point source spectral irradiance. This enabled the calculation of spectral irradiance spatial distribution formed by point source modules at the focal plane. Combining with relative radiometric calibration ratios measured by highly uniform integrating sphere light sources, we derived spatial distribution coefficients for the relative radiometric calibration of point source modules. Post-launch secondary correction of calibration coefficients ultimately formed the on-orbit relative radiometric calibration and non-uniformity correction algorithm.Laboratory measurements of irradiance distribution at various imaging positions under different driving currents confirmed system stability with ratio distribution relative standard deviation ≤0.05%, verifying current-independent performance stability. Algorithm validation employed image data with integration times from 8 ms to 16 ms to simulate surface reflectance variations. Post-correction results demonstrated significant non-uniformity improvement quantified by generalized noise assessment: non-uniformity decreased from 1.75% to 0.97% at 8 ms, 1.55% to 0.89% at 12 ms, and 2.28% to 1.28% at 16 ms. A multi-LED combination calibration strategy with energy-level matching enables wide-spectrum multi-level radiometric energy input through established energy-current relationships.System performance tests revealed exceptional stability LED source on-off cycling stability over 30 minutes exceeded 0.08%, with post-irradiation energy variation across all levels below 1%, fully meeting on-orbit stability requirements. This methodology provides an effective solution for real-time radiometric correction of spaceborne payloads, demonstrating significant application value for ensuring reliable on-orbit instrument operation.

Erbium-doped laser materials directly emit mid-infrared laser radiation in the 2.7~3 μm spectral region through the 4I11/2→4I13/2 transition of Er3+ ions. This spectral region, proximal to the strong water absorption peak, is valuable in medical surgery, environmental pollutant monitoring, and optical parametric oscillators, and thus has recently attracted considerable attentions. However, the self-terminating effect of Er3+ ions severely limits the oscillation efficiency of mid-infrared lasers. By enhancing the energy transfer upconversion process through increasing the doping concentration of Er3+ ions, the self-terminating effect can be effectively suppressed by simultaneously decreasing the population in the lower energy level and increasing the population in the upper energy level. However, the thermal effects of the crystal induced by high doping concentrations limit the further improvement of laser performance. Dual-wavelength cascade output can be achieved through 4I11/2→4I13/2 and 4I13/2→4I15/2 transitions of Er3+. The cascade method effectively suppresses the self-terminating effect at lower doping concentrations, reduces crystal thermal effects, and improves the efficiency of mid-infrared lasers, thus avoiding the problems caused by high doping concentrations while suppressing the self-terminating effect.A theoretical analysis of the cascade laser generation mechanism was conducted. Based on the Er3? energy level structure and rate equation model, the influence of the doping concentration of Er∶YSGG crystal on the cascade emission was investigated. Thermal simulations were performed on Er∶YSGG with different doping concentrations. It was observed that the axial temperature of the crystal increased significantly with increasing doping concentration. Consequently, 1 at.% and 5 at.% doped Er∶YSGG crystals were used for cascade experiments. The crystal length used in the experiment was 10 mm, with a diameter of 3 mm. A 969 nm wavelength-stabilized fiber-coupled laser diode was employed as the pump source. To reduce the mid-infrared laser output threshold, the output coupler was designed with high reflectivity at 1.4~1.7 μm and 2.5% transmission at 2.7~3.0 μm, with a total cavity length of 28 mm.The cascaded output mechanism has demonstrated significant efficacy in enhancing the performance of near-infrared lasers. Under 50 Hz pulsed pumping conditions, the maximum near-infrared pulse energy of 1 at.% doped Er∶YSGG reached 3.47 mJ, with a slope efficiency of 5.5%, significantly surpassing non-cascaded operation. An output energy of up to 2.82 mJ was obtained at a pumping frequency of 100 Hz, with a slope efficiency of 6.2% in the cascade state. Furthermore, continuous-wave pumped cascaded output of Er∶YSGG crystal was realized for the first time, achieving a near-infrared output power of 448 mW, and a slope efficiency of 8.2%. The cascaded output spectrum was measured, revealing a single peak at 1 645 nm in the near-infrared region. The mid-infrared spectrum exhibited multiple peaks with increasing pump power, and the variations in the four mid-infrared emission wavelengths at 2 788 nm, 2 829 nm, 2 870 nm, and 2 925 nm were analyzed based on small-signal gain and Boltzmann factor distributions.Simultaneous emission of 1 645 nm near-infrared and~3 μm mid-infrared laser radiation was achieved. The near-infrared output significantly enhanced the slope efficiency of the mid-infrared laser. Experimental results demonstrate advantages of 1.0 at.% doped Er∶YSGG crystal under room temperature cascade operation. The concurrent near-infrared and mid-infrared laser output provides a viable approach for developing efficient, continuous-wave, room-temperature mid-infrared lasers. Further enhancement of the room-temperature mid-infrared output performance of Er∶YSGG crystals is anticipated through optimization of crystal dimensions, doping concentrations, and resonator design, facilitated by the cascade approach. The results have potential applications in medical surgery, environmental monitoring, and free-space communication.

Compared to traditional 2D image detection, 3D point cloud detection provides several distinct and unique advantages that make it highly beneficial and widely applicable in various fields, including autonomous driving, drone navigation, and robotic obstacle avoidance. These advantages include an enhanced ability to resist interference caused by light sources and shadows, the ability to directly obtain 3D environmental data, and the capability to detect objects at significantly greater distances. A key issue with voxel-based methods is the loss of fine-grained details during the voxel encoding process. This loss makes it difficult for the model to capture important features in the point cloud data. Additionally, the feature extraction process may be insufficient, leading to the model not fully exploiting the available information in the point cloud. To address these challenges, a 3D object detection algorithm has been proposed that combines multi-attention voxel encoding with a composite backbone network. The central idea of this approach is to introduce a multi-attention voxel encoder during the voxel encoding process, which includes three types of attention mechanisms: point attention, channel attention, and voxel attention. These mechanisms work together to enable the model to focus on the most important features in the point cloud, allowing it to extract more detailed and specific local and channel features while minimizing the loss of fine-grained information.In addition to the improvements in the voxel encoding process, the backbone network of the model is structured as a composite network, which integrates multiple different networks. A feature fusion module is used to connect these networks laterally, allowing them to share feature information. This cross-network integration improves the model’s ability to understand and represent complex scenes, which in turn enhances the accuracy of object detection. Furthermore, to tackle the issue of class imbalance in the dataset, the classification loss function is modified by introducing a perturbation term. This modification fine-tunes the loss values for each category, allowing the model to better handle the imbalance in the number of samples from each class, which ultimately leads to improved classification accuracy. The proposed algorithm consists of four key components: voxel encoding, the backbone network, the neck, and the detection head. The process begins with the input of unordered 3D point cloud data, which is then spatially divided into pillars. After this, the point cloud data undergoes voxel encoding and multi-attention voxel encoding processing, transforming it into 2D pseudo-images that can be processed more easily by the network. The backbone network is made up of two identical down-sampling structures, each consisting of convolutional layers, normalization, and activation functions. These structures are linked by a feature fusion mechanism, known as attention feature fusion, which combines the features from both networks to enhance feature extraction and reduce information loss. The neck of the model utilizes a feature pyramid network, which is responsible for merging multi-scale feature information, allowing the model to better handle objects of different sizes and shapes. The detection head of the model further improves the classification process by incorporating a perturbation term into the original classification loss function. This addition allows for vertical adjustments of the polynomial coefficients in the loss function, which reduces the number of iterations needed to achieve the same loss values as the original model. As a result, the model is able to achieve higher classification accuracy more efficiently, requiring fewer training iterations.The effectiveness of this algorithm has been validated through experiments conducted on the KITTI public dataset. The results of these experiments show that the proposed method significantly improves the average detection accuracy for cars, cyclists, and pedestrians, with increases of 1.77%, 1.53%, and 7.68%, respectively. These results indicate that the model is able to maintain real-time performance while substantially improving detection accuracy. The improvements are particularly notable for pedestrians, who are often difficult to detect due to their smaller size and less distinctive features. Overall, the proposed method has proven to be a highly effective solution for improving 3D object detection performance in autonomous systems, especially in complex and dynamic real-world environments.

Nowadays, perovskite materials have shown significant application potential in the fields of up-conversion luminescence and various optoelectronic applications due to their excellent physical and chemical properties. Therefore, Rare Earth (RE) doped perovskite materials have attracted considerable interest to further improve the accuracy of the temperature sensors. In this work, a series of Tm3+, Ho3+, Yb3+ doped double perovskite Gd2MgTiO6 were successfully prepared by high temperature solid phase method. Raw material was fully ground in agate mortar for 40 min and then the grounded samples were transferred to a muffle furnace and sintered at 1 250 ℃ for 4 h. The structure, microstructure and luminescence principle of the samples were studied by X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM) and up-conversion luminescence spectroscopy. The phosphor with the best luminescence intensity was selected to investigate temperature sensing characteristics based on the Fluorescence Intensity Ratio (FIR) technique which is widely used to develop new temperature sensing materials.The experimental results show that the diffraction spectrum of the undoped sample in the XRD pattern is consistent with the literature report. The position and shape of the diffraction peak after doping are basically the same, indicating that the rare earth ions enter the double perovskite matrix without significantly changing the crystal structure. In the SEM image, it can be seen that the sample is uniform and the surface is smooth. The doping of rare earth elements does not affect the morphology of the material. The Gd2MgTiO6:1%Tm3+/0.5%Ho3+/5%Yb3+ sample exhibits the best luminescence intensity under 980 nm infrared laser excitation. The emission peaks of the sample are located at 477 nm, 545 nm, 654 nm, 759 nm and 800 nm, corresponding to 1G4→3H6(Tm3+), 5F4, 5S2→5I8(Ho3+), 5F5→5I8(Ho3+), 5F4, 5S2→5I8(Ho3+) and 3H4→3H6(Tm3+) energy level transitions, respectively. Compared with the double-doped samples, the luminescence intensity of the triple-doped Tm3+, Ho3+, Yb3+ is 6 times higher than that of the double-doped Ho3+, Yb3+, because Tm3+ ions can transfer energy to Ho3+ ions through the cross-relaxation process. When the excited state energy level of Tm3+ matches the energy level of Ho3+, Tm3+ can transfer energy to Ho3+ through a non-radiative process, resulting in enhanced upconversion luminescence of Ho3+. The emission of Gd2MgTiO6∶Tm3+/Ho3+/Yb3+ samples in blue (477 nm), green (545 nm) and red (654 nm, 759 nm) bands is caused by three-photon, two-photon and two-photon excitation processes.Based on the FIR technique, the temperature sensing properties of I545/I700, I654/I700, I477/I700, I800/I700 non-thermal coupling energy level pairs were studied in the temperature range of 293~573 K. The phonon energy of the matrix material increases which could make the non-radiative relaxation process more effective and promote the energy transfer from Tm3+ to Ho3+. With the increase of temperature, the intensity of each emission peak of the sample gradually increases. The luminescence intensity of the 700 nm band gradually increases. This luminescence phenomenon is mainly due to the electron transition from the 3F2, 3 energy level of the Tm3+ ion to the 3H6 energy level. The increase in temperature causes the particles to be thermally excited to a higher energy level. The number of particles at 3F2, 3 increases, and the non-radiative relaxation phenomenon makes the I700 emission intensity gradually increase in the temperature range of 293~573 K. The relative sensitivity of the sample reaches a maximum value of 1.72% K-1 at 573 K. At the same time, the maximum relative sensitivity of the synthesized phosphor and other rare earth doped luminescent phosphors is listed. The sensitivity of the fluorescent material prepared in this paper is higher than that of most optical temperature measuring materials, indicating that the sample has application potential in the field of temperature sensing technology.

Electromagnetic Induction Transparency (EIT) effect, originating from quantum destructive interference among diverse excitation pathways, enables remarkable suppression of light absorption within specific narrow spectral regions in originally opaque media. This effect endows materials with distinctive transparent windows in their transmission spectra, accompanied by significant dispersion properties. Such characteristics have propelled EIT to the forefront of numerous research endeavors, particularly in optical sensing, slow light effect exploitation, and nonlinear optical device fabrication. Nevertheless, the realization of quantum EIT effect is shackled by stringent experimental requisites, including stable lasers and ultra-low temperature environments. These constraints have severely circumscribed its practical applicability and further evolution.This study design an all-dielectric metamaterial for the terahertz band, predicated on the bright-dark mode coupling approach. The unit cell of this metamaterial is composed of two hollow cylinders fabricated from LiTaO?, an ionic crystal renowned for its pronounced polarization response in the terahertz regime. The material's complex permittivity adheres to the Lorentzian dispersion model, with parameters meticulously calibrated based on established literature.To comprehensively investigate the optical properties and EIT-like effect of the designed metamaterial, a finite element algorithm is employed. This numerical technique enables the accurate modeling of the electromagnetic response of the metamaterial under diverse conditions. Through this approach, the transmission spectrum is simulated to identify resonant frequencies and assess the quality factor of resonances. The multi-pole scattering power is calculated by decomposing the induced current in Cartesian coordinates, facilitating the quantification of the contribution of each multi-pole mode to the overall resonance. Additionally, the spatial distribution of the electromagnetic field at resonant frequencies is analyzed to visualize the field patterns and corroborate the excitation of specific multi-pole modes.The transmission spectrum of the metamaterial reveals a highly sharp toroidal dipole resonance in the vicinity of 0.879 THz. Employing the second moment method, the linewidth of this resonance is determined to be approximately 0.019 GHz, corresponding to an exceptionally high Q value of 4.63×104. In contrast, a magnetic dipole resonance centered around 1.068 THz exhibits a relatively broader linewidth of 19.8 GHz and a moderate Q value of 53.8. The multi-pole scattering power analysis further attests that the toroidal dipole dominates the resonance at 0.848 THz, with its scattering power surpassing that of other multi-poles by a factor of 2.5. Conversely, the magnetic dipole assumes primacy at 1.068 THz. The spatial distribution of the electromagnetic field at resonant frequencies vividly depicts the circular displacement current oscillations within the hollow cylinders, unequivocally validating the excitation of the toroidal dipole.By judiciously adjusting the center distance between the two cylinders within the range from 20 μm to 40 μm, an EIT-like effect is successfully realized in the proximity of 1.102 THz. The Q value of the EIT-like transmission peak attains its maximum when d=29 μm. The multi-pole scattering power analysis in the EIT-like state divulges that both the toroidal dipole and magnetic dipole are vigorously excited in the frequency regions adjacent to the transmission dip. In the vicinity of the EIT-like transmission peak, however, their excitation is conspicuously suppressed, thereby corroborating the hypothesis that the EIT-like effect is engendered by the coupling between these two dipoles.The metamaterial's potential as a refractive index sensor is investigated by subjecting it to varying environmental refractive indices ranging from 1.0 to 1.3. Even minute variations in the refractive index precipitate a significant redshift in the EIT peak. The sensor's sensitivity is quantified as approximately 151 GHz/RIU within the refractive index range of 1.0~1.2. The Figure of Merit (FoM), calculated as the ratio of sensitivity to the transmission peak linewidth, is determined to be around 6.8.In summary, this research has culminated in the successful design and demonstration of a terahertz all-dielectric metamaterial with an EIT-like effect. The unit cell structure, consisting of two hollow LiTaO? cylinders, has been shown to support the excitation of a toroidal dipole resonance with an unprecedentedly high Q value of approximately 4.63×104, in conjunction with a magnetic dipole resonance. The physical mechanism underpinning the EIT-like effect has been unequivocally attributed to the coupling between the toroidal dipole and magnetic dipole modes.The metamaterial's remarkable sensing performance, characterized by a high sensitivity of 151 GHz/RIU and a FoM of approximately 6.8, positions it as a highly promising candidate for applications in environmental monitoring and on-site biochemical detection. Future research directions may focus on further optimizing the metamaterial's structure to enhance its performance and exploring additional application scenarios in emerging fields such as terahertz imaging and communication.

Photodetectors are crucial optoelectronic devices that facilitate the conversion of optical signals into electrical responses. They can be broadly categorized into two types based on their operational mode: externally powered and self-powered devices. Cobalt(Ⅱ, Ⅲ) oxide (Co3O4) has garnered significant attention as a promising semiconductor material for optoelectronic applications owing to its non-toxic nature, abundance on Earth, and dual optical band gaps. Graphene, with its exceptional optical, electrical, and mechanical properties, along with a large specific surface area, presents itself as an ideal candidate for use in doping materials. Bismuth vanadate (BiVO4), an N-type semiconductor, is recognized for its excellent stability under harsh conditions and has attracted widespread research interest in areas such as photoelectrochemical (PEC) water splitting, energy storage, dye degradation, and photocatalysis, due to its non-toxic, abundant, and cost-effective characteristics. By doping graphene to modify the surface of Co3O4 and constructing a heterojunction via the PN semiconductor principle, a novel approach for self-powered photodetectors is proposed. In this study, Co3O4 thin films and Co3O4/graphene(G) composite films, with varying graphene molar ratios, were successfully fabricated on fluorine-doped tin oxide (FTO) conductive glass using a hydrothermal method. The optimal graphene doping ratio for Co3O4/G films was determined, and BiVO4 was subsequently spin-coated onto the surface of the Co3O4/G films, resulting in the fabrication of (Co3O4/G)@BiVO4 composite thin-film-based photodetectors. Photodetector performance was evaluated using an electrochemical workstation (CHI760E) and a xenon lamp (CEL-S500) to simulate sunlight and measure the photogenerated current. Raman spectroscopy was employed to confirm the presence of graphene and evaluate its defect levels. The microstructure of the samples was analyzed using Field-Emission Scanning Electron Microscopy (FESEM) and transmission electron microscopy (TEM). The phase and chemical bonding states of the samples were investigated by X-ray diffraction (XRD) and X-ray Photoelectron Spectroscopy (XPS), while optical absorption properties were characterized using UV-Vis spectrophotometry (UV-3600). The results indicated that the Co3O4/G film with a graphene doping ratio of 1∶2 exhibited the most stable photogenerated current, with minimal decay, and its photocurrent was 10.8 times greater than that of pure Co3O4 films. Furthermore, the morphology of the (Co3O4/G) film transitioned from a uniform, compact grass-like structure to a graphene-like network. BiVO4, prepared by spin-coating, formed a block-like coating on the Co3O4 nanorods. The (Co3O4/G)@BiVO4 composite film exhibited enhanced photocurrent and light absorption properties, with a photocurrent 6.3 times higher than that of Co3O4/G films. The device demonstrated a responsivity of 2.52 mA·W-1 and a detectivity of 2.693×1012 Jones. This composite film structure can provide an ideal research idea for the preparation of simple, non-toxic and harmless, and miniaturized high-performance self-powered photodetectors.

With the increasing level of confrontation between precision-guided weapons, the advantages of dual-mode composite guidance with high hit accuracy are becoming more. This paper designs a common aperture laser/visible light dual-band composite photoelectric guidance component. The common aperture optical structure can reduce the system size, visible imaging subsystem can obtain high-definition images of the target, and the laser receiving subsystem can perform high-precision positioning measurements on the center of the laser spot reflected by target. We first use geometric optics method to calculate the focal length of the each optical subsystem, and then use optical simulation software of Zemax for design and optimization. The calculated focal length of the visible light imaging subsystem is about 80 mm and that of the laser receiving subsystem is about 17.7 mm. The result of design and optimization of Zemax is that the common aperture laser/visible light dual-band composite photoelectric guidance component consists of eight lenses, one filter and one splitting cube prism. The visible light imaging subsystem consists of a total of six lenses and one cube prism. The laser receiving subsystem consists of five lenses, one cube prism, and one filter. The two subsystems share three lenses and one cube prism. The entire system has an incident light aperture of 26.7 mm and an F number of 3. We also used Zemax optical simulation software to analyze the MTF curves and point plots of the visible light imaging system in the meridional and sagittal directions at the diffraction limit, 0°, 1.999 6°, 2.777 3°, 3.338 4° and 4° tilt angles at temperature of 20 ℃, -40 ℃ and 60 ℃, as well as the distortion plots of light at wavelengths of 0.480 0 μm, 0.546 1 μm, 0.643 8 μm, 0.450 0 μm and 0.750 0 μm, and the sum and difference ratio curve of laser spot energy distribution via deflection angle for quadrant detector based on the laser spot energy distribution with deflection angle 0°, 0.5°, 1°, 1.3° and 5°. All calculation results meet the design requirements. The entire system has identify vehicle targets of 4.6 m×2.3 m at a working distance of 5 km, and building targets of 10 m×0 m at a working distance of 15 km; the laser receiving system has a full receiving field of view of ±5°, a linear receiving field view of ±1.3°, and an interception distance longer than 8 km.Based on the distribution of optical paths, combined with the requirements of optical processing and mechanical structure processing for the dimensional accuracy of eccentricity, tilt, center thickness, and air gap of each group of lenses, this paper finally designs a common aperture laser/visible light dual-band composite photoelectric guidance component with good imaging quality. Visible light imaging subsystem lens are mainly composed of a main barrel, six lenses, a cubic prism, five pressure rings, and four spacers. Laser receiving subsystem lenses are mainly composed of a main barrel, five lenses, a cubic prism, a filter, two pressure rings, and three spacers. The overall optical system length is 90mm and the width is 44.5mm. After adding the mechanical structure, the volume is 93.6 mm×68 mm×50 mm and weight is 227 g. We used collimator, microscopes, and focusing screen to measure the physical focal length of the common aperture laser/visible light dual-band composite photoelectric guidance component. The final measurement results showed that the focal length of the visible light imaging subsystem is 78.25 mm and the focal length of the laser receiving subsystem is 19.20 mm. We take visible spectrum photo of buildings within 3 km using the common aperture laser/visible light dual-band composite photoelectric guidance component. The size of the text on the sign on the building is basically the same as the size of the vehicle, and this optical system can clearly recognize the content of the sign text on the building, obviously is able to identify the vehicle target. We conducted field of view calibration tests on the common aperture laser/visible light dual-band composite photoelectric guidance component using an electric turntable and a 1 064 nm solid-state pulse laser at the micro watt level. Within the range of ± 1.3° yaw angle of the turntable, the pre calibration yaw angle and turntable yaw angle are linearly related. The system is simple and compact, easy to assemble and adjust, suitable for the trend of miniaturization of seekers, has very high practical value.

This study is dedicated to elevating the quality of reconstructed spectral data cubes in rocket exhaust plume monitoring through the development of a novel reconstruction framework grounded in low-rank and convolutional sparse constraints. Traditional spectral imaging systems encounter significant limitations in real-time monitoring. This is primarily because they rely on time and spatial domain scanning, which is time-consuming and restricts their real-time applicability. Compressed Sensing (CS) based snapshot spectral imaging presents a promising alternative. It allows for the recovery of high-dimensional data from under-sampled measurements, thus circumventing the need for extensive scanning. However, existing CS reconstruction algorithms grapple with the challenge of balancing accuracy and efficiency. In particular, under severely ill-posed conditions, they struggle to preserve the spectral-spatial fidelity of the reconstructed data. To address these limitations, this research proposes a method that integrates low-rank and convolutional sparse constraints to enhance the quality of the recovered spectral data cubes. The innovation of this article lies in the decomposition of the reconstruction task of compressed sensing spectral acquisition systems into a joint solution of low-frequency and high-frequency parts. By introducing a convolutional sparse coding framework and low-rank constraints, the reconstruction accuracy is significantly improved. This method not only overcomes the problems associated with traditional low-rank constraints when performing image rank minimization constraints within local regions but also better preserves the structural features of the image through the global processing characteristics of convolutional sparse coding. In the framework of convolutional sparse coding, the reconstruction task of compressed sensing spectral acquisition systems is decomposed into the superposition of two distinct parts: a low-frequency smooth main structure part and a high-frequency texture detail part. For the reconstruction of the high-frequency texture detail part, a convolutional sparse coding framework is proposed. An ??2, 1 norm constraint is imposed on the sparse feature maps corresponding to the convolutional dictionary. This constraint ensures prior knowledge about the spectral dimension of the spectral data, thereby significantly improving the accuracy of the spectral dimension in the reconstructed data. Regarding the reconstruction of the low-frequency smooth main structure part, the use of global convolutional sparse coding is proposed. A convolutional dictionary specifically trained for the low-frequency part is combined with this approach, and a nuclear norm constraint is imposed on the convolutional feature maps. Specifically, this article commences by conducting a comprehensive analysis of the imaging process of compressed sensing spectral imaging systems. The Alternating Direction Method of Multipliers (ADMM) is employed to solve the reconstruction equation in a piece-by-piece manner, facilitating the partial reconstruction of compressed sensing spectral imaging systems. Experimental results clearly demonstrate that the proposed method achieves high reconstruction accuracy in both the spatial and spectral dimensions. When compared with traditional iterative optimization methods and those combined with deep learning, the proposed method outperforms them in terms of key indicators such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Spectral Angle Mapper (SAM). Notably, in terms of the reconstruction accuracy of the spectral dimension, the proposed method can effectively preserve the detailed information in the spectral data, enhancing the spectral consistency of the reconstructed data. The component decomposition strategy effectively resolves the information loss issues in traditional local low-rank approximations while maintaining global structural coherence. By synergistically combining spectral correlation modeling through nuclear norm constraints with translation-invariant texture representation via convolutional sparsity, the method attains state-of-the-art performance in preserving both spectral discriminability and spatial detail. The adaptability of the framework to snapshot spectral imagers has practical significance for real-time engine monitoring. Since accurate spectral reconstruction directly affects the reliability of fault diagnosis, this framework provides a valuable engineering solution. Future research could explore dynamic dictionary learning for different combustion conditions and the co-design of hardware and algorithms for onboard implementation. This research not only advances the theory of CS spectral reconstruction but also offers a practical engineering solution for aerospace propulsion health monitoring systems.