In order to overcome the complexity of the space environment and meet the space operating requirements of the space engineering, a micro space-borne camera optical system of 25-megapixel wide-spectrum confocal imaging was developed. This system adapted to the harsh environment of satellite launch and orbit operation, which had the advantages of impact shock resistance, high temperature and differential radiation resistance in space, small size and light weight. The designed system can be clearly imaged in the spectrum of 450 nm?800 nm, of which the focal length is 181 mm, the entrance pupil diameter is 45 mm, the field angle is 10.4o, the edge relative illumination is 0.81. The modulation transfer function (MTF) is 0.57 at 55 lp/mm and 0.33 at 110 lp/mm, the distortion is 1.2%, the quality of lens is 622 g, the overall dimension is Φ58.3 mm×117 mm and the anti-radiation performance is more than 5 krad. Through the simulation and optimization of the temperature adaptability, the user can perform the thermal vacuum test of ?30℃~+70℃ optical lens normally. This system is successfully applied to the accompanying satellite camera of Tiangong-2 space lab, the obtained images are clear and stable, which plays an important role in the space remote sensing experiment observation.

Target location is one of the typical functions and tasks of the photoelectric turret, and its location accuracy is also one of the operational technical indicators of the turret. This problem was analyzed from the angle of theory and simulation. Firstly, the coordinate systems used in target location and their mutual transformation relations were analyzed. Secondly, the algorithms flow of photoelectric turret optic axis inversion, active target location and passive target location were given. Thirdly, considering the possible sources of random errors in location, the algorithms were simulated and verified, and the location results with or without mean-filtering were evaluated. Lastly, the location accuracy was calculated by the Monte-Carlo method. The results show that: 1) active location is better than passive location (under the assumption of simulation, the longitude is increased by about 1 time); 2) after mean-filtering, the location accuracy is improved (about 15 time); 3) the typical passive location accuracy of about 18 km distance is about 39.4 m under the condition of 80% confidence; 4) the analysis results of statistical histogram show that the radius of CEP at 80% confidence of location accuracy and the change results of the maximum error distance with different parameters such as aircraft position, attitude and turret optic axis, etc (reflected in the target-carrier distance).

In the process of image generation and sensing, the image is often disturbed by the noise, which will increase the difficulty to the subsequent image processing and even bring the huge economic losses to some production activities. Combined the advantages of stationary wavelet transform with convolutional neural network, an effective image denoising algorithm was proposed. In the training stage, the proposed algorithm was adapted to decompose the image with a stationary wavelet of scale 1, and the high and low frequency components were input into the four designed residual networks for training respectively. In the test phase, the inverse wavelet transform was used to obtain the final predicted image. The experimental results show that when the level of Gaussian white noise reaches σ = 50, the mean of peak signal to noise ratio (PSNR) and structural similarity index method (SSIM) of the denoised image can reach 28.37 dB and 0.808 0. This algorithm can effectively remove the Gauss white noise and natural noise in visible image, and the noise generated by remote sensing image in the sensing process. Moreover, the proposed algorithm can preserve the edge and texture details of the image while removing the image noise.

Accurate removal of the skeletal burrs is the most critical step in the extraction of interference fringe skeleton, which can be applied in laser interference fringe detection. A burrs removal algorithm of interference fringe skeleton based on skeleton features is proposed,which includes the acquisition of feature points of skeleton and the tracking of the eight-neighborhood linked list. First, the pixel points are scanned one by one to obtain the four feature points of the skeleton: endpoints, nodes, glitch points, and backbone points. Then, the algorithm of eight neighborhood linked list based on feature points is used to extract all the glitch points and backbone points, and the difference operation was performed based on nodes to remove the burrs. Finally, the processed image is iterated until the interference fringe skeleton burrs are completely removed. The OpenCV machine vision algorithm was used to simulate the burrs image removal, the results were verified by 1 000 pieces of burrs images, and the correct rate of burrs removal is 94%. Compared with the traditional scheme, the proposed algorithm has a higher pertinence, retains the backbone of the skeleton, and removes the remaining burrs, which has a broad application prospect in interference fringe detection.

In order to overcome the shortcomings of the existing methods for measuring plume velocity produced by solid rocket motor propellant combustion, a tunable diode laser absorption spectroscopy (TDLAS) technique was proposed to measure plume velocity. By measuring the characteristics of a single absorption line of hydrogen peroxide molecule in combustion products at 1 392 nm, the spectral frequency shift and molecular velocity were established according to Doppler effect. The relationship was used to obtain the airflow velocity, which solved the problem that the contact measurement method could interfere with the plume flow field and the inhomogeneity of tracer particles in the traditional non-contact measurement, and the effective experimental data are obtained. Futhermore, through the analysis and processing of the experimental data, the plume flow velocity of the engine was obtained.

Based on rainbow Airy theory, the sensitivity of the main Airy peak position and Ripple frequency to the refractive index and diameter was studied. Combined with empirical mode decomposition for rainbow signal, Airy signal and Ripple component were extracted from rainbow distribution. According to the first-order rainbow distribution of droplet, an inversion algorithm was designed to calculate the refractive index and diameter of droplet simultaneously, which with powerful feature of noise immunity. The first-order rainbow distributions from water and ethanol droplets with different concentrations were measured in the experiment. The results show that the refractive index and diameter measurement accuracy of the droplet by rainbow-empirical mode decomposition method are 10?4 and 1% respectively.

The light barrier is an important part for the headlamp in the modern projected automobile to adjust the light distribution performance of headlights. The focus method to adjust the position of the light barrier can effectively suppress the chromatism and gradient on the cut-off line, which has the functions of improving driver’s visual comfort and driving safety. An optical design of complex curved surfaces LED automotive headlamp based on light barrier with extended reflectors was adopted, the NX10.0 was used to build the structure model and LucidShape software was used for simulation, the maximum illumination value is 51.9 lx. The light distribution differences of optical design on light barrier with extended reflectors and optical design on conventional light barrier, as well as the comparison of two data under the road lighting standard proposed by Geely Automobile in 2017 were studied. The results show that the light barrier with extended reflectors can effectively improve the basic numerical of each point in the third light distribution zone and the visual brightness at 40 m~50 m ahead of the road, which are beneficial to the road ahead and pedestrians.

Based on the optical fiber laser hydrophone of distributed feedback (DFB), a new demodulation algorithm of single differential division (PGC-SDD) was proposed for the deficiency of traditional phase generated carrier (PGC) demodulation scheme. Compared with the demodulation algorithms of traditional differential cross-multiplication (PGC-DCM) and the Arctan, the improved algorithm could reach the demodulation results with reduced amount of operation steps and computation, in which only the signals in one of the two channels needed differential calculation. The PGC-SDD algorithm could effectively suppress the light intensity disturbance caused by environment and the harmonic distortion caused by modulation depth, which made the demodulation results closer to the signal to be tested. The simulation and experimental results show that the signal-to-noise ratio of PGC-SDD algorithm can reach to 17.5 dB, higher than 12.4 dB of PGC-DCM algorithm and 13.9 dB of PGC-Arctan algorithm.

In the deep working environment, the hydrostatic pressure can cause the reflection center wavelength of distributed feedback (DFB) fiber laser hydrophone drifting out of the multiplexing window of demodulation system, and the target sound pressure signal cannot be demodulated by the water receiver; different hydrostatic pressures also cause changes in the sensitivity of the distributed feedback fiber laser hydrophone, which affects the consistency of the fiber laser hydrophone array. Based on the theory of electric sound analogy, a distributed feedback fiber laser hydrophone probe with static pressure compensation was proposed. The sound pressure transfer function of the structure was established, and the influence of various structural parameters on the transfer function was analyzed, which provided a theoretical basis for the flattening design of the frequency response for distributed feedback fiber laser hydrophone probe. The resistant static pressure distributed feedback fiber laser hydrophone sample was processed and tested. The fluctuation range of sound pressure sensitivity was no more than ±2.4 dB in the frequency range of 0 kHz～10 kHz, and the wavelength drift of the exiting laser center within 2.3 MPa was no more than 0.06 nm. The results are of guiding significance for the research of deep water fiber laser hydrophone.

In order to improve the ability of driverless vehicles to identify objects in the surrounding environment at night, a three-dimensional target detection method for driverless vehicles based on multi-view channel fusion network was proposed. The idea of multi-sensor fusion was adopted, and the target detection was carried out by adding laser radar point cloud on the basis of infrared image. By encoding the laser radar point cloud into a bird’s-eye view form and a front view form, as well as forming a multi-view channel with the infrared night vision image, the information of each channel was fused and complemented, thereby improving the ability of the driverless vehicle to recognize surrounding objects at night. The infrared image and the laser radar point cloud were used as the input of the network. The network accurately detected the position of the target and the category by the feature extraction layer, the regional suggestion layer and the channel fusion layer. The method can improve the object recognition ability of driverless vehicles at night, the accuracy rate in the laboratory test data reaches 90%, and the time consumption of 0.43 fps also basically meets the practical application requirements.

In order to improve the quality of low-light night vision images, a color transfer algorithm based on image segmentation and local brightness adjustment was proposed. The simple linear iterative clustering was combined with K-means clustering to segment the low-light image, and the color component of matching reference image was transmitted to the sub-region of target image by using the uniformity for the brightness of each sub-region and reference image in the YCbCr color space. The contrast value in the texture feature of the target image was taken as the coefficient to adjust the brightness value of the sub-region of the target image, perform the color space conversion and display the color transfer results. A low-light image imaging system was built, and the low-light image segmentation and color transfer were completed. The results show that the improved segmentation algorithm separates different scenes in the image, and the peak signal-to-noise mean of the obtained color low-light image reaches 12.048 dB, which is 2.63 dB higher than the Welsh algorithm.

The method for tomographic computer generated cylindrical holography of three-dimensional object was researched: the diffracted wavefront on the holographic surface was the superposition of convolution between the cylindrical cross sections of three-dimensional object with different depths and the corresponding point spread functions, and the computer generated hologram could be obtained by recording the interference patterns from the diffracted object wavefront and the reference. The 360° view of the object could be observed from the reconstructed holograms. Firstly, the tomographic computer generated cylindrical holography model of three-dimensional object was built, and the conditions of system point spread function and sampling interval in different directions were derived. Secondly, the impact on the spatial frequency and the system transfer function by the radii and the wavelengths of different cylindrical cross sections was analyzed from both theory and experiments, and the peak signal to noise ratio as well as the mean square error were adopted to evaluate the quality of the reconstructed holograms. Finally, the tomographic computer generated cylindrical holography was used to encode the three-dimensional earth model, which represented the information of different observation angles and depths. The simulation results show that the proposed method has wide applications for 360° full field display of the ordinary three-dimensional objects.

Aiming at the problem that the cracked cell in the solar cell module eventually causes the whole cell to break and affect the power generation of the whole component, a method for detecting cracked defects of battery components using convolutional neural network network (CNN), is proposed based on the screening and positioning of the photoluminescence (PL) image of the battery component. The basic idea is to obtain the image of the battery component by using the PL detection technology first, then pre-process the image, filter and locate the target area based on the clustering method, and finally use three convolutional neural network models to detect the defect of the battery, and compare the accuracy. A large number of experimental results verify that the above method can accurately detect the cracking defects of solar cell modules.

TensorRT is a high-performance deep learning and inference platform. It includes a deep learning and inference optimizer as well as runtime that provides low latency and high throughput for deep learning and inference applications. An example of using TensorRT to quickly build computational pipelines to implement a typical application for performing intelligent video analysis with TensorRT was presented. This example demonstrated four concurrent video streams that used an on-chip decoder for decoding, on-chip scalar for video scaling, and GPU computing. For simplicity of presentation, only one channel used NVIDIA TensorRT to perform object identification and generate bounding boxes around the identified objects. This example also used video converter functions for various format conversions, EGLImage to demonstrate buffer sharing and image display. Finally, the GPU card V100 was used to test the TensorRT acceleration performance of ResNet network. The results show that TensorRT can improve the throughput by about 15 times.

Multi-dimensional information acquisition by single detector is the future direction of photoelectric detection. Aiming at the problem that the energy and the polarization information could not be considered in target detection, a new pixel array structure with polarization-low level light integrated function was proposed. The integration of polarization and low level light detection was realized in electronic multiplying charge coupled device (EMCCD) by introducing the white light channel and the reduced polarization channel. The experimental results show that under the low level light condition, the high sensitivity of the new detector is held, and the imaging quality at low illumination is almost not decay. In polarization mode, the white light channel and two polarization angle enable the detector to obtain enough polarization information, so as to realize the polarization detection of the target. Synchronous acquisition of high sensitivity imaging detection and polarization information detection by this method is realized, and the reconstruction of the detection mode can be realized by the algorithm processing.

In order to solve the problem that the data of the inner wall detection system of the body tube is large and the data processing is complicated, the parameter of the body tube is difficult to obtain, and the data processing method of the inner wall detection system of the gun barrel is proposed. On the basis of completing the construction of the gun barrel detection system, aiming at the special structure of the body tube and the characteristics of the data collected by the laser displacement sensor, the method of reducing the inner wall contour of the body tube and the method of separating the line data are proposed. The parameters of the inner wall of the body tube are obtained based on the least squares method. And the error of the system is analyzed and corrected, which improves the accuracy of the system. The experimental results show that the radial error of the system is less than 0.01mm, the error of the abnormal data correction is less than 0.01 mm, and the data processing method is correct, which effectively solves the problem of obtaining the inner wall parameters of the gun barrel.

The astronomical telescope of large field of view, low-cost and high-performance is the focus of current research and development. According to the aberration balance principle, the oblique incident reflective Schmidt correction plate equation was derived based on the normal incident Schmidt correction plate. For an optical system with the focal length of 1 700 mm, the imaging field angle of 4°, the operating wavelength of 0.4 μm to 0.9 μm, and the F number of 4.25, the Schmidt correction plate equation was solved and further optimized by using ZEMAX software as the initial structural parameter. The design results show that, the modulation transfer function (MTF) of this system is more than 0.35 at the Nyquist frequency of 100 lp/mm in the full field of view, and the distortion is less than 2.5 %, which indicates that the imaging quality is close to the diffraction limit. After the optimized design, the maximum deviation between Schmidt correction plate and near spherical surface is 0.005 mm, and the special compensator can complete the surface shape high-precision detection combined with the interferometer. The design of the Schmidt system can provide references for the development of astronomical telescopes with large field of view and wide spectrum.

Ambient temperature is one of the main factors affecting the imaging quality of aspheric optical systems. The analysis method of thermal optical properties was used to analyze the thermal optical properties of an airborne camera aspheric optical system. The finite element method was used to analyze the thermal deformation of the camera optical system structure, and the rigid body displacement of the mirror surface was removed. The surface data was input into the optical software program for Zernike polynomial fitting, and the fitting results were imported into the optical design software to evaluate the imaging performance of aspheric optical system. The results show that the analysis method of thermal optical properties can effectively simulate the practical working environment of the aspheric optical system, and predict the influence of ambient temperature on the imaging quality of optical system, which have guiding significance for the optical system design.

For the optical system with long focal length, the large relative aperture means that the imaging brightness is better, but it is difficult to correct with the variation of aperture edge aberration as well. The refractive system was used to reduce the total length of the optical system. Based on the reflection structure, the optical system was composed of two sets of correcting lens before and after, and the telephoto optical system with large relative aperture and short total length was effectively designed. The operating band of the optical system is the visible band, the focal length is 1 000 mm, the F number is 2.1, the telephoto ratio is 0.52, the optical total length is less than the focal length, the obstruction ratio is 45%, the full-field MTF is greater than 0.3 at the space frequency 80 lp/mm, and the image plane diameter is 11 mm. All the lenses of this optical system are spherical mirrors, and the system is composed of two reflectors and seven transmission mirrors, with compact structure and good imaging quality. The tolerance analysis of telephoto objective lens shows that the tolerance is good.

Aiming at the problems of low energy utilization and environmental sensitivity of passive light source adopted in high-speed projectile flight parameter measurement system, a combined active laser shadow photography system was studied. Based on the theoretical analysis of the system, the space position model and two velocity measurement models of high-speed projectile were established. The experimental platform of laser shadow photography system was built and the rationality of the system design was tested. The test results show that the both models can measure the flight speed of projectile, and the comparison error between the two methods is small.The mean square deviation of the projectile space position in X-axis is 0.795 mm and that in Z-axis is 0.496 mm. Compared with the results of paper target, the deviation degree is within 1 mm, and the system can measure the flight parameters of the projectile.

A coaxial alignment method of zoom system based on aberration feature analysis was proposed. Aiming at the alignment for a certain type of 20 times zoom optical system, the coaxiality of the optical elements in each lens group of the zoom system was improved by using the optical centering processing technology on the basis of analyzing the alignment requirements of the zoom optical system. The Zygo interferometer was used to detect the aberration feature distribution of the zoom system in the long and short focal positions. With the help of CodeV software, the tolerance sensitivity distribution of each optical element in the system was simulated to determine the sensitive optical elements which generated aberration effect. The final imaging quality of the optical system was adjusted on the center deviation measuring instrument. In addition, a coaxiality debugging device of the zoom optical system was designed to precisely measure the parallelism between the axis of the mechanical inner hole of the zoom camera main lens barrel and the linear guideway, so as to ensure the coaxial accuracy of the optical axis of the moving group. The results show that the imaging quality of the zoom system is obviously improved and the consistency of the image plane is guaranteed. The transfer function value of the long and short focal axes is better than the technical requirement value 0.45 and 0.55 respectively, and the transfer function value of the external long and short focal axes 0.7 field of view is superior to the technical requirement value of 0.25 and 0.35 respectively, which realizes the high-precision alignment and verifies the reliability of the method.

In order to solve the problem that when installing the slit lamp in the rotary drum, the human eye has an uncertainty of calculating the center distance of the cross-hair in the drum image, a digital adjustment technique based on the improved probabilistic Hough transform was proposed. By preprocessing the original image, improving the probability Hough line detection, and formulating the line intersection screening rules, the high-precision drum image cross-hair straight line detection and cross-hair center distance calculation were realized. Experiments show that the improved probability Hough transform can accurately detect the crosshairs in the image and fit the excess straight line 100% into 4 straight lines; and the intersection point screening rule can be used to accurately screen out 2 effective intersection points. The adjustment technology can realize digital assembly of rotary drum and the digital information can be used to remind the workers of the adjustment of drums better or, which can meet the actual needs of the drum production line.

Composite films are widely applied for their superior properties compared with the single component films. Taking the composite thoughts of protective, catalytic, electrical, optical along with mechanical properties as the points of penetration, the methods of mechanical properties were expounded through the membrane layer composite doping in order to enhance the corrosion resistance of alloy, promote the lubrication friction property, improve the conductivity of membrane layer, control the refractive index of optical thin films and the spectral absorption, as well as increase the hardness, tensile strength. The related frontier achievements at home and abroad are briefly introduced, and the future development of composite films is prospected, which provide reference for the research in related fields.

The gas detection method based on tunable semiconductor laser absorption spectroscopy has the advantages of non-contact, high resolution and sensitivity. However, the on-line detection of industrial gas by laser spectroscopy is susceptible to temperature changes, resulting in an increase in concentration measurement deviation. Study on ammonia, the mechanism of temperature effect on the absorption spectroscopy characteristics was explored, the experimental platform for ammonia laser detection at various temperatures (298 K to 323 K) was built, and the gas absorbance-temperature correlation method was proposed to correct the concentration inversion results. The results show that, the ratio of total partition function rQ is the dominant parameter affecting the spectral line intensity of ammonia at the same concentration. The spectral line intensity decreases with the increase of temperature due to the linear decrease of rQ with the rise of temperature. When the ambient temperature of ammonia reaches 323 K, the inversion value is 3.13%, and the relative error between the inversion value and the standard concentration value is as high as 37.4%. After correction, the relative error between the inversion value and the standard concentration value is in the range of 0.2%~1.4%.

Terahertz time-domain spectroscopy (THz-TDS) was used to detected the trace crude oil mixed into the sands away from wells due to the high sensitivity to polar substance. A linear model was established between the trace crude oil concentration and the terahertz attenuation coefficient. Effective-medium theory was introduced to confirm the linear relationship between the terahertz dielectric constant and the crude oil content below 200 ppm. The research suggests that terahertz time-domain spectroscopy can be an effective method for detecting the trace crude oil leakage in sands.

The high-speed measurement and reconstruction of dynamic liquid film 3D morphology is very important for the industrial process optimization in energy and power fields. Based on the Fourier transform profilometry, the simulation was performed for both slow and fast varying surfaces. The influencing factors of reconstructive accuracy on object surface morphology such as surface height change rate, environment random noise and fringe frequency were studied. And according to the simulation results, the optimal system parameters were obtained, high-speed 3D structured light measurement system was established and the surface morphology of dynamic liquid film was measured. The experimental results show that, as the liquid film flows downward along the vertical wall surface, the liquid film thickness first increases and then decreases, and the average error of the height direction is 0.1 mm. This work demonstrates that Fourier transform profilometry can be well applied in the high-speed measurement of dynamic liquid film.