With the development of marine economy and ocean exploration, the demand about the detection methods with high accuracy and efficiency becomes urgent. As a technology which combined the spectral technology and imaging technology, the hyperspectral imaging technology can bridge the gap of underwater high-precision detection demand to some degree, so the researchers begin to develop it and use it in the underwater detection. The background of the development of underwater hyperspectral imaging technology is first introduced, then the different methods of hyperspectral imaging technology and the components of underwater hyperspectral detection system are summarized. The research progress of underwater hyperspectral imaging detecting technology is summarized after that. The challenges which underwater hyperspectral imaging technology faces are pointed out, and the expectation of the underwater hyperspectral imaging technology is given.

The excitation of bioluminescence by different flow regimes generated with a Couette chamber was examined using the dinoflagellates Pyrocystis noctiluca, Electron-Multiplying CCD was used to detect at a fixed position. In this paper, luminescence characteristics of P. noctiluca were qualitatively studied under different flow field stimulation. An open water tank containing a certain concentration of P. Noctiluca was used to simulate the sea surface environment, and a water model with a propeller was used to make counterclockwise circular motion. The image data obtained from EMCCD was processed and analyzed based on the single-scale Retinex algorithm, so as to maintain the maximization of optical wake characteristics and suppress the influence of the surface background. After image enhancement, the wake of target motion can be clearly observed to rotate counterclockwise, which preliminarily proves that the visualization of target motion trajectory in water can be realized by using the excitation light phenomenon of Marine organisms. Therefore, biologically stimulated light can be used as a new technique and method to detect moving objects in the ocean.

The characteristics of laser downlink cross-media transmission through atmosphere, rough sea surface and seawater under three typical weather conditions (clear day, stratus and cirrus) are analyzed by using Monte Carlo simulation method. The results show that, the angular distributions of the light propagating through atmosphere are different under different weather conditions. The divergence is small under clear day condition, which is serious under stratus condition and obvious under cirrus condition. The difference between the spot sizes under clear day and stratus conditions is small, while the size under cirrus condition is about 15% larger. The laser shows different characteristics propagating through atmosphere and sea surface under different weather conditions. The angular divergence is larger and increases with wind speed under clear day condition. It is smaller and decreases with wind speed under stratus condition. The angular divergence changes complicatedly under cirrus condition. Under different weather conditions, there are similar characteristics of light propagating through atmosphere-surface-sea as well as differences. The characteristics of the angular distributions are different in shallow water, while similar in deep water. The spot size changes little with water depth under clear day and stratus conditions, while it decreases as the depth increases under cirrus condition. In addition, the spot center energy attenuation with depth is different under different weather conditions. Under clear day condition the average rate of change is about -0.410 dB/m, while under stratus and cirrus conditions about -0.426 dB/m and -0.413 dB/m respectively. The results could offer a reference for blue-green laser cross-media communication link budget, system design and parameter optimization.

In order to study laser array in the whole atmosphere turbulence coherent imaging ability, on the basis of all phase method such as closure, the atmospheric refractive index structure constant of HV57 model is used to calibrate the atmospheric coherence length, through the local wind speed and light road height to calculate the experimental parameters, such as local equivalent whole layer of the atmospheric turbulence level transmission distance. A laser coherent emission array with a diameter of 1.5 m equivalent aperture is developed, and the imaging experiment of a 1.2 km distant object is carried out. The imaging resolution reaches sub-arcsecond level. It is verified that the laser coherent array imaging system can suppress the influence of atmospheric turbulence and provide theoretical and technical support for the development of solid laser coherent array imaging system in the future.

From September to October 2019, the coherent Doppler lidar observed and retrieved PM2.5, PM10 particulate concentrations in a joint observation campaign conducted at Shiyan(113.9°E, 22.7°N) of Shenzhen. Statistical regression analyzed the retrieval with lidar backscattering intensity and synchronized hybrid ambient real-time particulate monitor measurements from different heights of meteorological tower. Correlation coefficients of PM2.5, PM10 particulate concentrations intercomparisons between lidar and monitor reach more than 0.8, and that of PM2.5 is better. Hygroscopic growth factor analysis shows large particles of 2.5 μm~10 μm at 120 m and 220 m heights may have stronger hygroscopicity, while it is opposite for those at 70m height. Particle concentration inversion and intercomparison prove that Doppler lidar can be used to observe particle concentration.

Based on the theory of Kolmogorov oceanic turbulence spectrum and quantum optics, the theoretical model that the spatial two-qubit photons entangled states prepared by parametric down-converted propagate through the Kolmogorov oceanic turbulence is constructed. The theoretical expressions for entanglement degradation of the spatial two-qubit photons entangled states in oceanic turbulence are obtained. Then, using the Wootters's concurrence, the influence of Kolmogorov oceanic turbulence on the spatial two-qubit photons entangled states with numerical simulation is analyzed. The results show that the parameters of the laboratory device which are prepared the spatial two-qubit entangled states will make a great impact on the entanglement. And the smaller separation of two signal (idler) apertures or the separation between the two signal and the idler apertures is, the higher fidelity of the spatial two-qubit photon entangled states is. The entanglement of spatial two-qubit states can well maintain in the salinity-induced oceanic turbulence when the rate of dissipation of mean-square temperature is small and the rate of dissipation of turbulent kinetic energy per unit mass of fluid is big with numerical calculation. These results have important significance for long distance underwater quantum communication via quantum entangled channel.

A multi-channel and time-resolved single-photon detection system was described, which is applied to high-sensitivity laser induced spectroscopy measurement. The photon detection system is based on a 32-channel linear array photomultiplier tube working at high-gain state. After amplifying and threshold discriminating, the single-photon pulse signal is converted to digital signal with LVDS level. Using the serial-to-parallel conversion hardcore resources of FPGA chip, the signal is sampled by a high-speed 800 MHz clock bilateral edge sampling. After the serial parallel conversion, the signal is logically parsed at a relatively low speed of 200 MHz clock frequency. The system realizes a 32-channel photon detection with 625 ps time resolution and 5 ns dead time on a single XC7K160T FPGA chip. Using this detection system, a 532 nm pulsed laser induced spectrum observation system based on grating spectroscopy was built. It realizes range-resoved laser spectral detection with 620 nm~720 nm spectral range and 3 nm spectral resolution.

Aiming at the test requierments of modulation transfer function for ultraviolet focal plane arrays, a system based on knife-edge scanning method is built to measure the modulation transfer function. Knife-edge is clearly imaged on the detector by a SW-Cassegrain reflective optical system and actual test shows that the optical system is near the diffraction-limitation in range of 0~100 lp/mm. An ultraviolet focal plane array device is tested by this system and the principle of maximum area under modulation transfer function curve is adopted to realize automatic focusing of knife edge image. The edge spread function curve is fitted by the model of three term Fermi function, which can effectively increase the fitting accuracy. Experimental results show that the measured modulation transfer function value of ultraviolet focal plane arrays is about 0.601, which is close to its theoretical limit. Meanwhile, the repeatability is good, and relative standard deviation of modulation transfer function whithin the Nyquist frequency range is about 1.8%. Uncertainty analysis indicate that maximum uncertainty of the equipment is about 6.2%, which can meet the needs of practical application.

When observing a target with a short-wave infrared microscope, the vertical size of the target is often limited by the depth of field of the lens. Aim at this problem, a multi-focus fusion method that can effectively expand the depth of field of microscope images is proposed. A large number of images with different focal planes is obtained by changing the object distance. Multiscale and local weighted variance is used to quantify the area definition. Then the focal plane mask is obtained. The morphological methods is used to optimize the mask boundary. Finally, focal plane regions are used to obtain the fusion image with complete details by weighted fusion. In the experiment, a microscope lens with a depth of field of about ten microns is used to obtain a high-quality image of a target with a vertical size of several hundred microns. The experimental results show that compared with other methods, the proposed method has certain advantages in terms of keeping the boundary of the focus plane and the details.

Aiming at the problem that the existing reversible information hiding has a low embedding rate and can not hide multiple different images at once. An optical reversible information hiding method for multiple images in ciphertext domain is proposed. Multiple images of different types and sizes are reconstructed by datagram. Multiple images can be reconstructed under certain noise without providing any information. The carrier image is encoded with asymmetric double random phase coding to obtain an encrypted image. Then the light intensity of the recombined multi-image is reduced by 1 000 times, and it cascade interference with the encrypted image to obtain the carrying secret image. Using cascaded vector decomposition, multiple images and encrypted images can be restored losslessly from encrypted images.The carrying secret image can be decrypted with the decryption key, or hidden key can lossless restore reassemble multiple images from carrying secret image. The carrier image can be restored lossless when both the decryption key and the hidden key are available.Realize completely reversible and separable.The experiment shows:When the embedding rate is 128 bits/pixel, the peak signal-to-noise ratio of the carrying secret image is greater than 32 dB. Hidden images and carriers can be restored lossless.Under clipping 1/2 or various noises of 0.2, the peak signal-to-noise ratio of each extracted and restored image is greater than 11 dB. Has good robustness. 32 bits/pixel can be embedded, extracted and decrypted within 3 s, which is highly efficient.

Due to uneven lighting at night and the complex environment, the current nighttime dehazing algorithm has problems such as light source diffusion, color distortion after dehazing, and bad image quality. According to the characteristics of haze images at night, this paper proposes a nighttime dehazing algorithm based on mixed filter light estimation and transmission optimization. Aiming at the inaccurate estimation of the ambient light at night, it is proposed to first perform side window filtering on the brightness image to ensure the illumination direction, and then use the guided filtering to refine the three channels as the local ambient light estimation; to solve the problem of light source diffusion caused by the application of dehazing at night, High-light area compensation is used to improve the transmission of the light source area. At the same time, for the uneven color and loss of details after dehazing, it is proposed to use guided filtering to correct the coarse transmission and then regularize the solution. Finally, the dehaze image is solved by the atmospheric scattering model. Experimental results show that the algorithm achieves good dehazing effects at night: in subjective evaluation, the algorithm can dehaze on the basis of preserving the original color of the image, improving image details and effectively suppressing light source diffusion; In terms of objective evaluation, all evaluation indicators of the algorithm have been greatly improved.

A single-frame on-axis digital hologram reconstruction method based on deep learning is proposed to suppress the zero-order and twin images in on-axis digital holography based on the powerful feature extraction capabilities. The U-Net is used to train and reconstruct different kinds on-axis holograms, including intensity and phase targets. The results show that the U-Net-based neural network can achieve high-precision reconstruction of the on-axis holograms. A set of on-axis holograms based on letters with different noise levels are generated to verify the robustness of the U-Net-based neural network. The results show that the U-Net-based neural network is robust to different targets and noise levels, and the structural similarity of the reconstruction results is better than 0.92.

To improve the accuracy of the commonly-used Extended Kalman Filter that omits the high-order terms of the derivation of the system equations, the second-order Adaptive Extended Kalman Filter (AEKF) was furtherly developed and Unscented Kalman Filter (UKF) was introduced based on two-compartment model. The results obtained by first-order AEKF, second-order AEKF and UKF based on numerical simulations and in vivo experiments were assessed and compared. The results demonstrate that the fluorescence pharmacokinetic parameter reconstruction based on first-order AEKF and second-order AEKF are similar, while the parameters obtained based on UKF are optimal in terms of quantitativeness and contrast-to-noise ratio, which is consistent with the theory that UKF should have higher accuracy because it does not ignore higher-order terms, and validate the feasibility and effectiveness of our proposed methods.

An in vivo method for measuring the depolarization characteristic of human tissues based on polarization-sensitive optical coherence tomography is proposed. Human finger nails are used as samples. Stokes vector and the macroscopic Mueller matrix measurement methods are adopted to obtain images of different depolarization parameters of nails. The advantages and disadvantages of these two methods are measured by comparing the images. In the images generated by the Mueller matrix-based method, the nail bed structure can be clearly distinguished. It has high contrast, but with showing less structure clearly. In the Stokes vector-based method, the depolarization uniformity values obtained by different irradiation polarized light at the same position are different. The 45° polarized light is selected to calculate the circular polarization of its backward reflected light which derives the structural information of the sample, including the nail plate, the nail bed and the dermis under the nail matrix. The circular polarization measurement method based on Stokes vector incident 45° polarized light has a small amount of calculation, and extracts the most structural information, and the imaging effect is the best.

In order to realize the high-resolution reconstruction of static target objects hidden behind the scattering medium and the reconstruction and real-time tracking of dynamic target objects, avoid the unpredictability of pseudo-thermal light sources generated by rotating ground glass and the impact of mechanical vibration on imaging quality, a spatial light modulator is applied to the speckle-related imaging system to generate the pseudo-thermal light field, enabling imaging of static and dynamic targets. The error rate of dynamic target tracking is within a very small range, the minimum can reach 0.38%. In the image reconstruction process, a phase recovery algorithm based on phase retrieval with the swept approximate message passing algorithm is adopted to improve the calculation speed and image robustness. Comparison of the quality of the pseudo-thermal light source and image reconstruction based on phase and amplitude modulation methods is conducted, the result shows that the quality of the pseudo-thermal light source based on phase modulation is better, as well as the corresponding reconstructed image resolution.

The existing non-invasive focusing technology has the problems of complex calculation, low anti-noise performance and slow optimization speed. Therefore, a wavefront shaping-based method is proposed. Based on the linear excitation of fluorescence, by maxmizing the linear combination of the intensity and contrast of the detected fluorescent speckle in an appropriate contribution, a phase-only spatial light modulator is used to optimize the input light so as to realize the focusing in the scattering medium. The experimental results show that the proposed method not only realizes deep focusing on a single particle in the scattering medium, but also has a fast convergence speed and is insensitive to noise. Moreover, it provides a new way for non-invasive deep imaging in complex medium and is expected to be used in life science.

Hydrogen diffusion is an important factor that causes additional absorption loss in optical fibers. The concave G.657 fiber with the inner cladding was chosen as the experimental fiber, and the structure and attenuation factors of the G.657 fiber were analyzed. The mechanism of deuterium gas to eliminate the hydrogen sensitivity of the fiber was explained, and the fiber hydrogen diffusion experiment was designed to carry out the deuterium gas treatment formula. By adjusting the two key parameters of deuterium concentration and processing time, the additional attenuation value of the fiber under different experimental conditions was obtained. The comparison experiment results and the tracking retest results show that 0.9% deuterium gas concentration and 80 h deuterium treatment time are suitable for reducing the hydrogen diffusion of G.657 optical fiber.

To address the shortcomings of piezoelectric and optoelectronic pulse wave sensors that are susceptible to interference in strong electromagnetic interference environments, a flexible material encapsulated Fibre Optic Grating (FBG) pulse wave sensor is proposed for measuring human radial artery pulse signals. The optimal package thickness and the position of the optical fibre in the substrate were first investigated using comsol finite element simulation software. Based on the simulation results, the sensor thickness of 5 mm and the optical fibre encapsulation at 1 mm from the lower surface of the substrate were optimised to combine the strain transfer effect and the flexibility of the sensor. The FBG flexible pulse wave sensor was fabricated and the human flexural artery pulses were collected from ten test subjects. The denoised signals were well preserved by a modified threshold wavelet method, and the signal-to-noise ratios were above 40, and the peak, tidal and repulse waveform recognition rates were 100%, 100% and 90% respectively. The results show that the flexible substrate FBG pulse wave sensor can effectively acquire and identify pulse wave signals, providing a theoretical basis and application support for further applications.

A Fabry-Perot Interferometer (FPI) was constructed by fusion splicing a Hollow Core Fiber (HCF) with a length of tens of microns at the end of a single-mode fiber and coating gelatin film at the free end of the HCF. Relative humidity measurement was achieved by detecting wavelength shift of the interference spectrum with humidity level changes. Experimental results show that high sensitivity of 192 pm/%RH has been obtained in the temperature range of 20 ℃ within the range of 20%~80%RH and the measurement accuracy and repeatability are quite good. In addition, the sensitivity of 173 pm/%RH and 194 pm/%RH were obtained by humidity increasing experiments at 15 ℃ and 25 ℃, respectively. In order to measure temperature has also been achieved by cascading a fiber Bragg grating sensor with the FPI sensor head. The optical fiber humidity sensor possesses several advantages including simple fabrication, high sensitivity, and temperature measurement. It has good potential in the field of humidity measurement.

In order to improve the measurement accuracy and repeatability of monocular vision camera, a closed-loop control method for the scanning mirror in laser scanning projection is proposed to improve the stability of the projected fringe position. The closed-loop control is carried out by using the feedback signal provided by the piezoresistive sensor integrated on the scanning mirror. At the same time, according to the temperature characteristic of piezoresistive sensor, a test system is designed to calibrate the relationship between piezoresistive output and temperature. By recording the piezoresistive feedback output value at different temperature, a table of the relationship between the feedback output and temperature is generated. In the temperature range from room temperature to 70 ℃, the scanning angle change of the scanning mirror decreases from 3.52° to 0.05°. By compensating the scanning angle of the scanning mirror, the 3D testing accuracy and the repeatability of the testing data of the monocular vision camera are greatly improved.

In order to monitor the micro-scale deformation of optical fiber material during industrial production, image wedge model is built and then a out-of-plane displacement measurement technique based on optical flow field is proposed. The experimental apparatus consists of an optical microscope with an industrial camera and a holder with a fixed image inverter and the deformation can be loaded by tilting the holder. First, 2D images before and after deformation are captured through microscope with CCD camera. Subsequently, the optical flow field between two images is obtained by Brox optical flow algorithm. Finally, according to the image wedge model, the out-of-plane displacement field between two images is extracted from the optical flow field and the results of experiment are analysed as well. The experimental results indicate that the absolute errors and the relative errors of the measurement by optical microscope with magnification of 50× are less than 0.1 μm and 1.5%, respectively. The displacement measurement can be completed by two consecutive frames which obtained by only one optical microscope with a industrial camera. The proposed method is appropriate for dynamic deformation monitoring and industrial in-situ detection and it neither need conversion of the images to frequency domain, also do not need phase envelope operation during the out-of-plane displacement extraction process. Due to the low errors, it has been applied to the industrial uniformity and micro-nano-scale deformation monitoring of optical fiber image transmission materials.

Aiming at the limitation of the current target tracking performance evaluation method of photoelectric theodolite, an improved video signal injection test method is proposed based on the advantages of traditional optical signal injection test method. Firstly, according to the planned trajectory of the target during the test, the spatial mapping relationship between the target and the optical axis of the theodolite is calculated, and the corresponding target image and background image are generated, which are projected to the photoelectric theodolite by the target simulator, so as to establish the database for target and background image respectively. Secondly, according to the spatial position of the target as well as the spatial mapping relationship between the target and the optical axis of the theodolite at each measurement time, the target image and the background image in the image database are called. The target image is blurred according to the relative motion speed of the target, and then fused with the scene image, and injected into the video processor of the theodolite, so as to realize the tracking performance test of the theodolite with no-delay signal injection. The experimental results show that, the proposed video signal injection test method can achieve the measurement accuracy of outfield evaluation. It not only makes full use of the advantages of traditional optical signal injection test method, but also effectively solves the problems of the target simulator's projection delay and the need to follow the theodolite. It provides a new and effective way for tracking performance evaluation of photoelectric theodolite.

Terahertz time domain spectroscopy and support vector machine algorithm are combined to study the defect identification method for multilayer bonded structures. On the one hand, the linear discriminant analysis method was used to reduce the dimension of 14 terahertz time-domain characteristic parameters extracted by the terahertz time-domain spectrum system, and the classification accuracy of normal region, debonding region and edge region in the adhesive layer of multi-layer bonded structure was improved by 20.3%. On the other hand, chaos particle swarm optimization was used to optimize the kernel function of support vector machine, and the classification accuracy of adhesive layer Ⅰ and Ⅱ increased by 18.92% and 9.85% respectively. Linear discriminant analysis based on constructed after parameter optimization of chaotic particle swarm optimization algorithm of support vector machine for multilayer glue joint structure characteristic imaging, the results show that this imaging method can effectively distinguish between sub area of the normal, defect region and edges region, compared with the traditional characteristics of terahertz single imaging technology promoted the debonding defect recognition rate of 50% above,The recognition rate of adhesive layer Ⅰ is 91% and that of adhesive layer Ⅱ is 92%, which greatly improves the recognition ability of debonding defects of multi-layer adhesive structure.

In the manufacturing process of high-resolution mobile phone lens consisted of glass and plastic material, many problems of assembly between glass lens and plastic lens group were existed such as sensitive to position change, low efficiency and low yield. The optimal proposal based on multi-point infinite conjugation modulation transfer function measurement was adopted to lens alignment and assembly in real-time feedback. As a critical imaging component of chart acquisition for modulation transfer function calculation, performance of multi-axis focusing lens played a significant role in guaranteeing measurement accuracy. The relationship between the design parameters of multi-axis focusing lens and the optical parameters of mobile phone lens to be assembled was established. Multi-configuration setting and aberration optimization of multi-axis focusing optical system were realized by ZEMAX simulation software. Simulation results showed that the size of focusing lens was compact, and both its tangential and sagittal modulation transfer function curves were nearly coincided with diffraction limit. The designed multi-axis focusing lens theoretically could be flexibly exploited to multi-point modulation transfer function measurement and precise assembly of high-resolution mobile phone lens up to 48 megapixel, which maximum full field of view angle reached 106°.

A high spectral-resolution solar Far-Ultravoilet/Ultraviolet (FUV/UV) spectrometer carried by high-altitude balloons to the bottom region of the near space was designed, which will historically fill the data gap in wavelength shorter than 280 nm that deposits engergy in this region. The solar FUV/UV spectrometer is based on improved Roland circle optics. In order to adapt to the requirements of high-altitude balloon platforms, it has a compact design for the optics to reduce the dimensions of the instrument, to suppress the spherical aberration of the optical system, and to improve the quality of the spectrum. The optical performance was simulated by the Zemax software to analyze the spectral resolution, wavelength coverage, and stray light suppression.

Aiming at the disadvantages of large-scale solar simulators in the past, such as complex structure, low energy utilization rate and poor uniformity, a design method of solar radiation simulation optical system with high energy utilization rate was proposed. The ellipsoid condenser was designed based on the luminous characteristics of the real xenon lamp. By adding a spherical reflector, the combined condenser system was used to improve the energy utilization rate. The optical integrator was designed according to the principle of pupil matching. The irradiation uniformity was improved by the edge elimination method. The simulation results suggest that the energy utilization rate of the combined condenser system is 21.2% higher than that of the single ellipsoid condenser. The irradiation uniformity of the edge elimination method is 4% higher than that of the edge compensation method. The experimental results suggest that the working distance is 20 m, irradiation surface diameter is Φ2 m, maximum irradiance is 1 363.1 W·m-2, non-uniformity is ±4.5%. It has realized large-area, high-energy utilization rate solar radiation simulation, which provided an advanced means for semi-physical simulation and testing of solar sensors in space field.

In order to realize the detection, identification of long-distance infrared radiation targets. Through the research and analysis of the model and theory of harmonic diffraction, free-form surface and off-axis three-mirror optical system, we propose a design scheme of off-axis triple-mirror dual-band infrared imaging optical system based on harmonic diffraction and free-form surface. By selecting the initial structure through the analysis of quantitative indicators and using Zemax macro language to control the process of system optimization, a dual-band (3~5 μm, 8~12 μm) infrared imaging optical system with a focal length of 1 200 mm and a diameter of 300 mm was designed. When the system is at a spatial frequency of 10 lp/mm, the value of the modulation transfer function in the 3~5 μm band is greater than 0.6, and the value of the modulation transfer function in the 8~12 μm band is greater than 0.45. The root-mean-square radius of the diffuse spot of the system's dual-band full field of view is less than 25 μm. The system realizes non-thermal sensitization in the working temperature range of -60℃?~?+60℃. The designed system not only has a simple structure, meets the requirements of various technical indicators, but also has high realizability, which is of great significance to the further development of the airborne dual-band infrared imaging optical system.

All process steps in the fabrication process affect the performance of silicon heterojunction solar cells. In this contribution, we optimize the performance of SHJ solar cells by step-by-step analysis including scanning electron microscope, reflectivity, quantum efficiency and minority carrier lifetime measurement. It is indicated that the optimum pyramid size for c-Si wafer passivation is about 6~9 μm. More than 5 ms of minority carrier lifetime was obtained by passivating silicon wafer with high quality intrinsic hydrogenated amorphous silicon (a-Si:H) film. Large band gap p-type a-SiCx:H was used as emitter layer alternative to p-type a-Si:H film, which will increase the photoresponse in the short wavelength range. A significant improvement of photoresponse in the long wavelength range was also improved by reducing the free carrier absorption of indium tin oxides. Based on this optimization silicon heterojunction solar cells with power conversion efficiencies exceeding 21.68% were prepared on c-Si wafers textured in alkaline solution.

In order to accurately identify the target of the long-distance ship type under the condition of airborne photon radar, the perform plane fitting, point cloud clustering, target extraction and other processing on the scene point cloud to obtain the ship point cloud firstly. Then from the hull point cloud, three-dimensional features such as volume, surface normal vector histogram, and deck target distribution are extracted to obtain a feature array. Finally, the random forest is used to discriminate and classify the extracted features to realize the accurate identification of the hull type. Experiments show that through multiple classification experiments on 13 types of ships, the average correct recognition rate is above 95%, effectively realizing ship type recognition

The guard ring structure on the silicon pixel sensor is beneficial to improve the high voltage withstand performance of the sensor. In order to evaluate the protection effect of the guard ring structure on the silicon pixel sensor, three kinds of guard ring structures are simulated and analyzed. Two-dimensional modeling of the three guard ring structures was carried out by Technology Computer Aided Design, and the I-V characteristics of the three guard ring structures were simulated using the electrical model built in the software. The existence of the current collecting ring can make the pixel withstand high voltage, and the unequal-spaced guard ring, The different space guard ring, the inner and outer equidistant Al suspension of the guard ring, and the multiple guard ring structures are beneficial to further increase the breakdown voltage of the sensor.