In order to meet the needs of geostationary orbit high-resolution imaging and deep space astronomical observation, and on the basis of the rapid development of rocket carrying capacity, the caliber of space-based remote sensing cameras in China has developed by leaps and strides, from 1 m caliber to 3 m caliber and above. Wiffletree support structure is widely used in mirror assemblies with diameter of 2 m and above because of its uniform distribution of support points and strong deformation adaptability. However, due to the large number of support points and complex structural connection links of Wiffletree support structure, the traditional method of mirror assembly and positioning can not meet the needs of fast and accurate assembly. A multi-point joint positioning method based on the bidirectional limit of the optical machine was proposed. By completing the positioning assembly of the Wiffletree structure and the bearing plate of the camera in advance, and adjusting the relative position of the mirror and the bearing plate, the pointing accuracy of the mirror optical axis could meet the requirements. Finally, micro adjustment was made for each support point to ensure uniform bonding gap and minimum assembly stress. The method was applied to the installation and adjustment of a 3 m mirror assembly. The verification results show that the method has low stress and high efficiency, and can meet the requirements of the installation and adjustment of Wiffletree support structure of large aperture mirror.

For the development of the tracking frame in the field of low-orbit satellite-terrestrial laser communication, the motion performance requirements of the tracking frame for low-orbit communication was firstly analyzed, and then the overall design of lightweight and miniaturization was carried out. For the commonly used double-row ball rotary table bearing, the rigidity calculation model and method of this kind of bearing were proposed based on Hertz theory, and the maximum deformation of the ball was 2.92 μm. The friction torque-temperature relationship of the bearing was studied, and the friction torque of the bearing was reduced to 8.9 Nm at ?30 ℃ by the optimum selection of clearance and grease. The simulation results of the tracking frame show that the maximum deformation of the tracking frame under the load of 180 kg does not exceed 0.02 mm, and the fundamental frequency is more than 20 Hz, which has good static and dynamic characteristics. The development of the prototype was finally completed. The accuracy test results show that the PV values of the azimuth and pitch axis system wobble error are ±2.8″ and ±4.1″, respectively. The satellite-to-ground digital transmission experiments clearly illustrate the feasibility of the proposed tracking frame, which has a certain reference value for the development of similar tracking frames.

In the manufacturing process of aluminum foil, gold foil, etc., there will be a certain proportion of defective products. The typical defect is a tiny void, and the size of the defect is very small, usually at the μm level, which cannot be directly judged by the human eyes. A non-contact system device for detecting tiny defects by combining the highly collimated narrow-band line light source and a large field of view receiving optical system was proposed. The system consisted of the laser light source, a Powell prism, and an aspheric lens to output a highly-collimated, highly-uniform, and large-sized linear laser beam. The linear array detector was placed on the defocus plane of the focusing lens, and the specific location of the defect could be determined by the position of the received light spot. According to the intensity of the light, the size of the defect aperture could be identified, and finally the location and classification of the defect could be achieved. The system realized the detection of large-size products and μm-level tiny defects, and has non-contact, fast, and accurate automatic detection capabilities, which effectively solves the problems of low efficiency, high difficulty, high missed detection rate, and high cost when manually detecting tiny defects.

Under the condition of airborne uncontrolled firepower, the parameter requirements which affect the tracking and aiming of moving targets were analyzed. Through the establishment of a stable control architecture model based on a fast mirror, many error factors such as optical image error, load vibration stress error, dynamic response lag error, friction torque interference error, mechanical structure error and electrical error affecting the aiming accuracy were specifically elaborated, and the optical image error, load vibration stress error and dynamic response lag error were compensated and simulated by Matlab/Simulink. The simulation results show that the closed-loop bandwidth, stable tracking accuracy, and timeliness performance of tracking and aiming conversion of the control system are good, which theoretically proves the effectiveness of this tracking and aiming method.

Aiming at the phenomenon of signal spectrum broadening and overlapping in short-range ranging of frequency-modulated continuous wave (FMCW) lidar, a multi-interference optical path ranging system based on FMCW was proposed. The signal frequency was reduced by improving the optical path structure to ensure the ranging accuracy. The problem of low signal-to-noise ratio (SNR) caused by spectrum overlapping was avoided, and the function of frequency reduction was realized, so that the low bandwidth detector could collect the high frequency interference signal. Measurement experiments were carried out for different distances from 0 to 30 m. The sampled data were processed by short-time Fourier transform to obtain high-precision micro-Doppler time-frequency graphs, and the frequency features were extracted and represented as corresponding distance information by calculation. The frequency resolution stability is improved by the structure from 28.3% to 4.4% in the range of 0 to 30 m, and the spatial resolution reaches ± 10 mm.

The accuracy of fast steering mirror (FSM) can be improved by introducing filters in laser radar system under long distance conditions, but the delay problem brought about will affect the control performance of FSM. Therefore, in order to maintain the response speed while improving the tracking accuracy, a delay compensated linear quadratic regulator optimal control method was designed. The method constructed a discrete state-space model based on the dynamics and electrical model of FSM, and the optimal control was carried out by introducing a predictive compensator that approximated the first-order inertial link in combination with a state-feedback apparatus and a quadratic performance index. Experimental results show that in comparison with proportion integration differentiation (PID) control, the delay compensated linear quadratic regulator optimal control method improves the step time by 61.83%, the tracking accuracy in the X-axis direction by 50.92%, and that in the Y-axis direction by 29.98%, which can make the scanning imaging system obtain more accurate and reliable observation data.

With the continuous improvement of the optoelectronic platform's requirements on the stability accuracy of the visual shaft, the traditional two-axis and four-frame stable structure no longer meets the need of the platform. Due to the limitations of rotating shafting, driving motor, vibration damping mechanism and its own structural strength and other factors, the larger the optical load, the more the rotating shafting, and the larger the volume and weight of the universal frame. Therefore, a new type of visual axis vibration damping stability frame design by placing 8 identical flexible vibration damping connection modes symmetrically on both sides of the center of mass of the inner ring frame of the photoelectric platform was proposed. The modal simulation and static analysis of the framework model were carried out by Ansys simulation software, and the stability performance of the design framework was evaluated effectively by NX and ADAMS co-simulation. Finally, the dynamic performance of the rigid connection and the flexible vibration damping connection frame was tested by the sweep frequency test system. The results show that the frame can provide vibration isolation and static support for the optical platform at low and medium frequency, and thus improving the stability and anti-interference ability of the optical platform.

The calibration process of converting from phase information into three-dimensional information is an essential and critical step, in fringe projection profilometry based on projectors and monocular cameras. The existing phase-height model only obtains the height information, while the triangular stereo model treats the projector as an inverse camera, and both the camera and the projector need to be calibrated. To this end, a simple and geometrically constraint-free new calibration method that combined the phase-height model and the triangular stereo model was proposed. Firstly, based on the Zhang Zhengyou's plane calibration method, the internal and external parameters of the camera were calibrated on a zero-height experimental optical displacement stage. Then, based on the height of the calibration plate, the experimental optical displacement information, and the absolute phase information of the corresponding height, the Levenberg-Marquardt algorithm was used to realize the solution of the phase-height nonlinear model. Finally, based on the camera imaging model and the height information, the x and y coordinates were calculated to obtain the three-dimensional geometric information. The proposed calibration method has no geometric constraints and avoids projector calibration, and the root mean square error (RMSE) is 0.082 2 mm after experimental evaluation.

The beam output by a laser is usually described by the Gaussian distribution. In order to improve the uniformity of the beam distribution and the energy utilization rate to meet actual needs, the beam needs to be shaped. Several shaping techniques were summarized and divided into two categories: static optical elements and dynamic optical elements. The research progress, basic principles, factors affecting the shaping effect and practical applications of aspheric lens group, birefringent lens group, diffractive optical element, microlens array and liquid crystal spatial light modulator shaping system were introduced, respectively. The advantages of liquid crystal spatial light modulator compared with other shaping methods were emphatically summarized, and its development prospects were prospected.

The green channel imbalance is a typical defect of Bayer image, which can be suppressed by correction of green balance algorithm, so as to reduce the maze noise of the color image after demosaicking, and improve the perception of image smooth area. With the development of image devices towards high resolution and large area array, the past green balance algorithm is difficult to meet the requirements of low delay and high pixel resolution. Therefore, a real-time green balance algorithm based on field programmable gate array (FPGA) was proposed, in which the smooth regions sensitive to the imbalance of green channels through local image gradients were extracted, thereby completing the green channel imbalance correction. While ensuring the resolution, the maze noises were removed, and the overall delay was less than the time of two lines of data transmission. The system design of front-end photosensitive camera, cache, image processing and final output image according to the features of FPGA hardware was completed, and the algorithm also could be compatible with other image processing functions. Experiments show that this algorithm can achieve microsecond-level delay while maintaining the resolution of pixel-level texture information of the image, and has excellent robustness for different levels of green channel imbalance defects and different scenarios.

In order to better mine and retain the details of the original image and improve the performance of image fusion, an infrared and visible image fusion algorithm based on chrominance loss in high exposure scenes was introduced, in which an auto-encoder structure was adopted, and it was composed of encoder and decoder. In the encoder part, an high exposure suppression module (EXBlock) for adaptive luminance detection was proposed to suppress the interference of high exposure to scene details, which could effectively balance the global contrast of images. In the decoder part, in order to realize the fusion of deep and shallow features, the multi-scale feature reconstruction method was used to realize the complementary and enhanced features in different scales. In the network training stage, in order to reduce the suppression of local color information in the high-exposure scene, the chrominance loss was designed and introduced to guide the fusion image to converge to the color information, so as to realize the color retention of the fusion image in the high-exposure scene. In order to verify the effectiveness of the algorithm, the subjective evaluation such as fusion result map and objective evaluation such as visual fidelity and mutual information was presented respectively in the experimental part. Compared with other algorithms, the proposed algorithm performs better and can better retain the chrominance information of the original images.

Point cloud registration is an important part of three-dimensional point cloud processing, but as a widely used iterative closest point (ICP) algorithm, it has high requirements on the initial pose and overlap rate of source and target point clouds, and is prone to local optimization. Therefore, a point cloud registration algorithm based on feature extraction and improved ICP was proposed, which mainly included two core parts: sample consensus initial alignment (SAC-IA) and ICP fine registration. Firstly, the incoming point cloud was preprocessed by OCtree accelerated voxel filtering to remove discrete points and streamline the point cloud. Then, the fast point feature histogram (FPFH) feature descriptor was calculated for rough registration of SAC-IA algorithm. Finally, based on the initial matrix obtained by rough registration, the normal vector constraint information was introduced to improve the ICP algorithm for fine registration. The algorithm was verified on the public data set. The results show that the proposed algorithm effectively improves the registration accuracy and efficiency compared with the classical ICP algorithm, and provides a feasible method for the subsequent 3D reconstruction and other aspects.

In order to further optimize the robust principal component analysis algorithm, a robust principal component analysis algorithm based on the orthogonal inexact Lagrange multiplier method was proposed. The collected image sequences of flat bottom hole defects were processed, and compared with the results of traditional image sequence processing algorithms including polynomial fitting, principal component analysis, independent component analysis and pulse phase method. The performance of each image sequence processing algorithm was quantitatively analyzed from evaluation indicators such as defect detection rate, peak signal-to-noise ratio (PSNR), root-mean-square error (RMSE) and entropy. The results show that each evaluation index of the robust principal component analysis algorithm based on the orthogonal inexact Lagrange multiplier method is optimal, in which the defect detection rate, PSNR, RMSE and entropy are optimized by 9.09%, 1.14%, 11.34% and 4.60% respectively compared with the suboptimal values.

To address issues with traditional 2D fingerprint acquisition technology, such as distortion and hygiene risks, a structured light 3D measurement system for high-precision fingerprint topography measurement was introduced. By optimizing the phase calculation, telecentric imaging 3D detection model, and fast calibration algorithms, the problems of low accuracy and complex calibration of small field of view 3D measurement system could be solved. A 3D fingerprint acquisition system was built in the experiment, achieving the collection of fingerprints and the measurement of 3D morphology. Under the conditions of a field of view of 39.9 mm×26.6 mm at a working distance of 138 mm, the measurement accuracy of the system reaches 30 μm. Compared with traditional methods, the errors are significantly reduced, and the fingerprint collection speed is better than 1.5 s. The relevant research work is conducive to solving the problems of high acquisition cost and low reconstruction accuracy of 3D fingerprints, providing technical support for further improving the measurement accuracy, recognition rate and real-time detection performance of fingerprints in police physical evidence examination and trace examination and identification.

The sub-aperture stitching interferometry method requires a certain overlapping region between adjacent sub-apertures. By removing the phase deviations of corresponding points in the overlapping regions, a complete surface map was stitched together. A weighted fusion stitching algorithm was proposed. By calculating the weights of each group of measurement results in the overlapping area, the splicing traces produced when the averaging method taking values in the overlapping area were improved, achieving a smoother splicing effect. Additionally, by combining the cylindrical two-position absolute measurement method, the influence of the reference mirror surface shape was eliminated, simplifying the process for rotation in the cylindrical three-position interferometric measurement method. The influence of misalignment errors on measurement results during sub-aperture adjustment was also reduced. The measurement data of sub-aperture were obtained at the confocal and focal line positions, and after stitching, the Chebyshev polynomial fitting was applied to derive the absolute surface shape of the entire cylindrical aperture. This method not only reduces dependence of cylindrical measurement on the reference mirror aperture and focal length, but also ensures measurement accuracy.

Ellipsometric measurement technology is a commonly used method for measuring parameters such as refractive index, extinction coefficient, and film thickness of thin film samples. An ellipsometric measurement device and method based on dual elastic optical differential frequency modulation was introduced. A measurement system based on elastic optical modulation was designed, which modulated the optical signal and loaded the ellipsometric parameter information into the optical carrier signal. The optical carrier signal was demodulated by using digital phaselocked technology to obtain the information of ellipsometric parameter, thereby achieving the measurement of thin film thickness. The sample to be tested was a circular silicon wafer coated with SiO2 film. The spectral ellipsometer (ESS01) produced by Quantuo Company in China measures the film thickness of 105.008 nm with a standard deviation of 0.731 nm, and the designed measurement device measures the film thickness of the sample to be 105.167 nm with a measurement repeatability of 0.091. Through comparative experiments with advanced domestic instruments, it is shown that the proposed method can achieve accurate and rapid measurement of ellipsometric parameters.

High-altitude balloon has attracted much attention because of its wide application fields, and its production process requires strict control of key quality links, material selection and welding of ball film. Compared with the material quality of the whole ball film, the welding part is more fragile, which is easy to cause the failure of the ball structure due to weld defects, resulting in serious flight accidents. Aiming at common basic weld defects, a method of weld defects detection based on polarization sampling and target contour marking was proposed. On the basis of applying polarized light to enhance the contrast of weld defect position in the image, the sample collection was carried out, and after image preprocessing and threshold segmentation, defect contour features were marked, and they were equivalent to defect features. The defect features were digitized, so as to detect different kinds of weld defects. Experimental results show that the method can detect three typical defects in weld production samples, with an accuracy of 91.37%. It has high processing efficiency, short response time, convenient data tracking, and obvious output results, which can replace traditional manual detection and realize the purpose of efficient detection of ball film weld defects.

A novel method for measuring the concentration of pure solutions based on the principle of light attenuation was developed. By measuring the transmitted light intensity of laser in sample solution and calculating the absorption coefficient according to Beer's law, the reliable measurement of low concentration solution was realized. At 532 nm wavelength, the concentration of a 0.000 1 mol·L?1 copper sulfate solution measured by this method shows that the systematic error is controlled within 10%, and the absorption coefficient of copper sulfate solution measured is consistent with that reported in other literatures. The results show that this method has high accuracy and reliability, and can meet the requirements of low concentration solution measurement in related fields.

Aiming at the problems of low time resolution and poor stability of traditional mechanical rotary compensator ellipsometry, a system model for solving Muller matrix based on differential evolution algorithm was proposed. On this basis, the microsecond-level measurement of full Mueller matrix was achieved by combining ultra-high speed photoelastic modulation. Firstly, by studying the working mode of the double driven photoelastic modulator (PEM), it was proved that in the pure traveling wave mode, fast and periodic rotation in the fast axis direction could be achieved. Then, the PEMs with driving frequencies of 60 kHz and 100 kHz were designed and fabricated, and a ultra-high speed ellipsometry model based on dual driving PEM was constructed. The light intensity was fitted through differential evolution algorithm, and a system model for solving Muller matrix was established. The optical period was found to be in microseconds, and the mean square error of fitting all elements of the sample Muller matrix was less than 0.001.

During the production and detection process of optical lens, it is inevitable that dust will adhere to the surface. Its appearance is similar to hole and pockmark defects, however it does not have a substantial impact on the performance of the lens. In order to solve this problem, a set of optical lens surface defect acquisition device was designed to collect high-contrast images and use an improved You Only Look Once version 7 (YOLOv7) network to achieve detection and classification. Firstly, the enhanced network incorporated depthwise separable convolutions within feature extraction module and improved the spatial pyramid pooling structure to decrease the parameter count and broaden the receptive area. Secondly, when integrated multi-channel fusion feature enhancement module into the backbone network, this module could enhance the accuracy of similar feature recognition and add the ability of information interaction between channels. Finally, the loss function was modified to normalized Gaussian Wasserstein Distance~~Smoothed Intersection over Union (NWD~~SIoU) to improve the network's attention to small targets. The experimental results show that the mean average precision for the final network's detection capabilities attains 92.9%, alongside a reduction of 11.8 MB in the model's weight and the frames per second reaches 62.4 f·s?1, which illustrates that the method can distinguish dust from pitting and sand defects quickly and effectively.

On the basis of the infrared radiation difference between ships and the sea surface, it is possible to use infrared detectors to identify target ships. The surface temperature field of the ships obtained by CFD software and the rough background sea surface field established by Blender software were used, and the transmittance in the oceanic atmospheric environment was calculated using the Modtran atmospheric software. Based on reverse Monte Carlo method, a thermal imaging algorithm was developed to obtain the 8 μm~12 μm infrared thermal images under different detection conditions. The results show that the infrared features of the ships in the thermal image are mainly concentrated in the chimney and power plant compartment, which is the main location for identifying the target. As the detection height increases, the difficulty of identifying the infrared features of the ships increases, and the detailed target features can be captured by adjusting the camera focal length.

Airborne target detection in the infrared has been commonly known as infrared search and tracking (IRST), and its detection targets are point source targets. The detection range is an important and systematic indicator of the IRST system, which directly affects the ability and effect of the system to perform tasks and is the important basis for evaluating the system performance. For high-altitude IRST system, the detection range is closely related to the each link in the entire imaging process, such as target, background, atmosphere, optical system, and imaging detector, etc. Starting from the infrared radiation characteristics of the aircraft target and combined the atmospheric transmission, the radiation intensity differences between the target and the background in the mid-wave and long-wave bands were given. Based on the point source target/background signal-to-noise ratio model, the comparison analysis of the mid-wave and long-wave performance was carried out, which had reference significance for the design and analysis of IRST systems.

In response to the measurement requirements of a large-scale arch dam chain 3D deformation measurement system, the technical specifications for inclination angle sensors were determined. A high-precision photoelectric inclination sensor scheme based on suspension wire was studied, which used incoherent collimated beams to form a measurement light field. The shadow spot of the suspension wire was amplified by an optical system to form a clear image and projected onto a linear charge coupled device (CCD). Based on the principle of spot position measurement, the displacement value of the shadow spot was obtained, and the tilt angle value of the measurement point was calculated in reverse. At the same time, in response to the long-term on-site operation requirements, a two-dimensional high-precision photoelectric inclination sensor structure was designed. Test results show that the resolution and range indicators meet the design requirements, achieving high-precision inclination measurement with a range of 3 093 seconds and an accuracy of 0.432 seconds.

In order to solve the problem of flexible encapsulation of fiber Bragg grating (FBG) sensor, a FBG sensor with flexible resin as encapsulation layer was designed. Firstly, the impact of flexible resin encapsulation was analyzed. Secondly, the sensors encapsulated with flexible resin, polydimethylsiloxane, silicone were fabricated, and the pressure conditions of the sensors were analyzed by using Ansys software. Finally, the load, temperature sensitivity and static performance of the fabricated flexible sensor were compared with the bare FBG. The experimental results show that the FBG sensor encapsulated with flexible resin has good response and sensing ability. The average sensitivity of load is 4.644 pm/N, the average sensitivity of temperature is 1.685 pm/℃, and the goodness of fit of sensitivity is above 99.8%. The repeatability is 0.008, the hysteresis is 0.019, and the linearity is 0.01, which are better than the polydimethylsiloxane, silicone encapsulated FBG sensor and bare FBG. The research can provide reference for flexible encapsulation of FBG.

In order to solve the existing problems for furnace monitoring, industrial endoscope imaging resolution is low and it is susceptible to dust, which affects the accuracy of temperature measurement. A dual-channel industrial endoscope with multiple spectrum in the visible band (0.48 μm~0.65 μm) and infrared band (3 μm~5 μm and 8 μm~12 μm) was designed and processed. It had a field of view of not less than 100°, a total aperture of less than 60 mm, and a total optical length of more than 1 000 mm. The modulation transfer functions (MTF) of the optical system in the visible band, mid-wave infrared band, and long-wave infrared band are greater than 0.2 at 180 lp/mm, 30 lp/mm, and 12.7 lp/mm, respectively. After passive heat dissipation treatment, the overall optical system was less affected by the thermal difference within the temperature range of 10 ℃ to 70 ℃. The tolerance analysis results show that the designed endoscope has high-resolution imaging quality in the range of 1 m to 5 m object distance, and can distinguish 4 mm line width at 5 m. The designed industrial endoscope has been tested and verified in the factory furnace, which can effectively solve the monitoring problems in the high-temperature and dusty environment of the furnace, and improve the operational safety and production efficiency of industrial furnaces and kilns.

The absorption cavity module is an important component of the electrical substitution cryogenic radiometer. When the cryogenic radiometer works, the absorption cavity and the thermal link jointly establish and maintain the thermal balance. According to the measurement principle of electrical substitution, the equilibrium state of light radiation heating and the equilibrium state of electrical substitution heating were established, respectively. In order to accurately describe the differences of temperature distribution and electrical substitution inequivalence between the two equilibrium states, the differential heating method and finite element simulation method were proposed to evaluate the inequivalence caused by heating paths. According to the results of simulation analysis and uncertainty analysis, the difference of temperature measured by heating path is 1.5×10?5 K, and the inequivalence of electrical substitution heating is 1.2×10?8 W.

In order to ensure the reliable communication using prefabricated optical fiber connectors in the precision control environment, and to meet the requirements of multi-mode and single-mode optical fibers in terms of transmission performance, a low-loss prefabricated optical fiber connection ferrule with guided positioning was designed. By analyzing the insertion loss, tilt loss and gap loss of the optical fiber ferrule, two design methods were adopted to reduce the ferrule connection loss: the structural design of the ferrule inclination angle and error compensation, and the matching of the end coating material. Precise positioning of ferrule pairing was achieved by aligning the MT sleeve with the PIN guide pin and guide hole. Based on the connection loss of 850 nm and 1 300 nm multi-mode and 1 310 nm and 1 550 nm single-mode optical fibers, tests were conducted by using the BINNA-MTS 3D interferometer. The results show that the average loss of prefabricated ferrule docking is about 0.3 dB and 0.24 dB, the maximum loss added value of mechanical properties and environmental test is 0.19 dB, and the maximum value is much less than 0.8 dB, which indicates that the end face and alignment performance of the fiber connection ferrule meet the requirements.