An imaging spectrometer with a varied line-space convex grating was designed and optimized in this study to realize a large relative aperture, high-resolution imaging, and light miniaturization, thus ensuring easy processing, assembly, and adjustment of the optical system based on a simple system structure. With a double-element system in a reflection concentric structure, the complexity and application limitations of the existing aberration-correction configurations were overcome. An aberration-correction design method was proposed based on the aberration theory of gratings, and a theoretical model of the geometric aberration of the system was briefly derived. The focal conditions of the system in both the meridional and sagittal directions were then analyzed in combination with the Rowland circles of the concentric configuration.This analysis established a mathematical model of the relationship between the astigmatism and line spacing of the convex grating. Finally, an anastigmatic design of the imaging spectrometer with a convex grating was realized using a global optimization algorithm. Design results show that the system can achieve high-quality imaging in the spectral range of 300~800 nm with a large relative aperture of F/2.7 and a resolution of 1.9 nm.In addition, approximately 201 hyperspectral channels are covered. The modulation transfer function for the whole spectral band is greater than 0.7, which meets the design requirements of the system. This study contributes significantly to the research of hyperspectral imaging spectrometers with aberration-correction convex gratings in terms of light weight and compactification.

For detecting mirror plane defects, fringe modulation is generally regarded as an effective method. To improve the accuracy of defect detection, this paper proposes a sector fringe-based modulation testing method. The proposed method is more compatible with the defect direction and suitable for many types of defects. A new quality map is used based on deformed fringe modulation and phase gradient deviation to identify the locations of defects. Eight groups of experiments are conducted and compared.Results show that the modulation and phase gradient quality map can suppress the selectivity of defect directions and the testing accuracy of defects can reach a minimum size of 0.07 mm. Compared with the traditional machine vision method and horizontal/vertical/circular fringe-based methods, the proposed sector fringe modulation method offers state-of-the-art sensitivity and higher accuracy.

Ultraviolet lasers with picosecond pulse widths can quickly break the molecular bonds of materials, reducing interaction time. At the same time, they have high temporal resolution. In order to obtain an ultraviolet laser with a picosecond pulse width, a solid fiber, hybrid-amplified 213 nm laser was designed. First, a 1 064 nm fiber laser with a repetition rate of 5 MHz, pulse width of 52 ps, and average power of 2.5 W was used as a seed source. After two stages of end pumped Nd∶YVO4 crystal amplification, a 1 064 nm laser with an average power of 10.5 W was obtained.This 1 064 nm laser and a 266 nm laser were then combined in a BBO crystal to produce a 213 nm UV laser.The final output is a 213 nm UV laser with a pulse width of 690 ps and an average power of 61 mW. The nonlinear conversion efficiency is 0.58%.

To reduce the edge effect of low energy outside the waist radius of a Gaussian beam when high-power laser equipment is operating and avoid the problems of edge and lap marks, it is necessary to design effective beam shaping. First, to reduce the complexity of the system and improve the utilization of the laser energy, a high-power Gaussian laser light source(average optical power of 500 W, pulse widths in 130-160 ns, wavelength of 1 064 nm, and power regulation range of 10%-100%) was optimized using the Galileo aspheric lens shaping system.Second, through the law of conservation of energy, the mapping relationship between incident and emitted light was obtained, and the structural parameters and aspheric coefficients of the system were optimized and simulated using optical software. The output energy uniformity of the flat-top beam at ejection distances of 20,25, and 30 mm was calculated. The results show that the energy uniformity of the flat-top region is greater than 78.0% when the focal depth of the system is between 20 and 30 mm. A laser cleaning experiment is conducted using a rusted low-carbon steel plate. The interaction between laser and material is uniform in the area of spot action, proving that the design can meet the requirements of high-power laser environments.

Silicon-organic hybrid photonic integration technology fully utilizes the advantages of large-scale integration of silicon photonics and the high electro-optic coefficient of organic polymer materials. This accords great application potential in high-performance integrated microwave photonics systems to it. This paper discusses a comprehensive optimal structural design and preliminary preparation of a silicon-organic hybrid integrated electro-optic modulator. The slit of the silicon slot waveguide is filled with electro-optic organic polymer. The optical field confinement factor reaches 0.32 when the structure of the slot waveguide is optimized. The optical wave field and radio-frequency electric field overlap considerably, thereby improving the electro-optic modulation efficiency. The tapered waveguide mode conversion structure is used to achieve low-loss coupling between the strip and slot waveguides with a acoupling efficiency of 99.55%. The frequency modulation response is addressed by analyzing the influence of the electrode length and width. The 3 dB bandwidth is optimized to 77 GHz, and the half-wave voltage-length product reaches 0.045 V·cm. The MASK used in manufacturing the modulator is designed based on the simulation results. Experiments involving the slot waveguide and the filling polymer in the slot waveguide are performed with good results.

The change of the equivalent refractive index of liquid crystal molecules in liquid crystal spatial light modulators has been used to achieve dynamic control of optical wavefronts, thus realizing zero position interference measurements of optical aspheric surfaces. However, the pixel pitch, filling ratio of the cell structure, and gray-level number directly affect the accuracy of wavefront control. This study is based on Fresnel diffraction theory and the discrete Fourier transform algorithm.The physical light field was analyzed using a numerical analysis software,VirtualLabTM.The influence of pixel-independent units on the reconstruction compensation wavefront precision of the liquid crystal spatial light modulator was investigated, and a simulation model of the wavefront propagating through the spatial light modulator to the surface to be measured was established.The theoretical algorithm's background error of wavefront reconstruction caused by the pixel structure factor was evaluated, and the multifactor coupling error of the image structure was determined by focusing on the sampling frequency of pixel size and the maximum space of thewavefront.For the frequency matching problem, a spatial light modulator was selected to meet the requirements of high-precision measurement according to the theoretical model of the reconstructed wavefront, thereby effectively reducing the cost of the dynamic detection system. Theoretical calculation shows that the factors that restrict the spatial light modulator in reconstructing the high-precision wavefront (root mean square error of 0.01λ) are closely related to the nonlinearity of the liquid crystal molecular response and the inconsistency of the substrate space.

To obtain the real colorful three-dimensional (3-D) point cloud data model of a target surface, a 3-D point cloud color restoration method is proposed. The method uses code points, based on the colorless 3-D point cloud and colorful images of an object, as the location medium. First, multiple code points were placed near the object and thecolorless 3-D point cloud of the object; thereafter,the information of code points was captured by a 3-D scanning device based on a structure laser line. A color camera was then used to capture color images of the target from different perspectives, and then, using canny edge detection and a subpixel localization algorithm, coding point coordinate values in the images were calculated. Later, an improved method of direct linear transformation was used to solve the camera pose matrix in the world coordinate system by coding point world coordinates and pixel coordinates.The corresponding relation between the point cloud and color image pixels was then set up by the camera imaging model. Finally, interpolation was used to calculate the color of the local point cloud, and thus, complete color fusion of the entire point cloud was achieved. In the experiments, the proposed method was adopted to restore the color of multiple point cloud models. The color space position deviation of the color 3-D point cloud is less than 0.6 mm, and the reconstruction efficiency is approximately 73 000 point/s.The color 3-D point cloud exhibits high similarity with the targets.Therefore, the method can achieve the goal of color 3-D point cloud reconstruction, and the achieved calculation accuracy and efficiency can satisfy user demands.

A method was proposed to eliminate the distortion of targets beyond the depth of field (DOF) in integral imaging systems and improve the overall effect of the reconstruction. First,the relationship of the diffuse spot diameter in the integral imaging system was analyzed through ray tracing, and the DOF range of the system with reference to the visual characteristics of the human eye was obtained. Then, in the case of a fixed collection device, the depth of the three-dimensional target beyond the DOF range was adjusted according to the actual depth value used as the reference value to obtain an element image using a look up table. Finally, the reconstructed image without distortion was obtained in the reconstruction stage. Experimental results demonstrate that the targets within the range of the acquisition DOF can be clearly reconstructed, thus verifying the efficacy of the acquisition DOF model. According to this model, the reconstructed image could be obtained using the depth adjustment method. Its color similarity with the two-dimensional color map increases from 67.057% to 94.507%. The SSIM increases from 54.002% to 84.510%, and the PSNR increases from 16.902 to 19.740. This method clearly eliminates the distortion in the reconstruction of objects beyond the DOF range and is suitable for resolution-first integral imaging display technology.

To achieve remote measurements of wind turbine vibration, a novel wind turbine state monitoring method was proposed in this study.Coherent Doppler lidar was used to detect low-frequency vibration of a wind turbine. The telemetry measurement, vibration spectrum data analysis, and all phenomena observed in the experiment were tested and studied. The measurement was divided into two parts.First, the top of the wind turbine was scanned via Doppler lidar at a distance of approximately 675 m, and vibration spectrum data were processed using discrete spectrum correction technology to extract the vibration state parameters.Second, a fixed target was measured, and the echo signal was processed using the same data processing method. Furthermore, system and measurement errors were evaluated. The results show that the coherent Doppler lidar is efficient at low-frequency vibration detection for a wind turbine with a vibration frequency of 0~2.5 Hz and a vibration speed of 0.1~3 m/s.

The ASRTU Microsatellite is a 12U CubeSat managed by the Harbin Institute of Technology and developed under the cooperation of the China-Russia University of Technology Alliance (ASRTU). This paper presents the design of an attitude control system for the ASRTU microsatellite mission. The control indicators of the ASRTU Microsatellite are first introduced and the main performance indicators of the attitude sensor and actuator are determined through index decomposition. Then, according to the configuration of the on-board sensor, a variety of positioning and control scheme designs are presented, and the switching logic between each control mode is described.The extended Kalman filter algorithm based on gyro and star sensors was used for satellite attitude determination, and the deviation quaternion and the deviation angular velocity feedback PD control method was used for attitude control. Simulation results indicate that the accuracy of attitude determination system when the star sensor is valid is better than 20". Moreover, the control accuracy and stability are better than 0.05° and 0.01 (°)/s, respectively.The results indicate that the proposed design meets the mission requirements of ASRTU CubeSat.

To address motion overshoot and low imaging speed of the hopping mode inconventional scanning ion-conductance microscopy (SICM), this paper proposes a high-speed scanning method based on a dual-stage nanopositioning system. A design scheme consisting of a long-range actuator operating along with a high-speed actuator is proposed, according to the need for the SICM measurement range in the Z direction.The pipet probe can approach the sample surface with a highspeed, without contact. Rhombus-type amplification and guiding mechanism are used as the basic configuration.The key parameters of the dual-stage nanopositioner are determined by an analytical modeling. The static and dynamic characteristics of the dual-stage nanopositioner are evaluated via finite-element analysis. The final prototype is processed and overshoot and imaging experiments are performed. Experimental results show that the designed Z direction nanopositioning stage driven by the dual actuator can significantly improve the probe approach speed to at least 500 nm/ms, which effectively improves the imaging efficiency of the hopping mode, without reducing the imaging stability.

Water-quality sensors feature a single measurement function and are greatly affected by ambient temperature. This paper proposes a multi-parameter water-quality sensor chip with temperature compensation. The sensor chip was manufactured using the micro-electro-mechanical system (MEMS) technology. The chip surface contained integrated pH, dissolved oxygen (DO), ammonia-nitrogen, and temperature sensors. To achieve temperature compensation, the chip was also designed with a sandwich-plate snake Pt resistance heater. At the same time, a microfluidic test chamber that was matched with the chip was designed to realize water-sample measurement. Finite-element steady-state thermal analysis was used to analyze the heat-transfer process of four heaters, and a reasonable chip-structure layout was established. Thereafter, the sensor chip was fabricated using the MEMS technology. The performance of the sensor chip was investigated using a laboratory-developed potentiostat test circuit. Experimental results show that the pH sensor has a high sensitivity of 0.288 mA/pH.The sensitivity of the temperature sensor is 0.949 Ω/℃ and that of the ammonia-nitrogen sensor was 0.113 9 mA/(mg·L-1). The sensitivity of the DO sensor is 2.22 μA/(mg·L-1) and that of the temperature sensor, which changes with respect to the heater power, is 0.312 6 ℃/mW. Compared with a single water-quality parameter sensor, the as-prepared sensor chip could simultaneously detect multiple parameters of a water sample and exhibited a good temperature-compensation effect. Moreover, the sensor chip was small in size, robust, and highly accurate.

High-speed electric spindle is a novel spindle that combines the drive and transmission units of the conventional spindle. The high-speed electric spindle is widely used in high-performance machining centers because of its high speed and high precision. The dynamic rotation characteristics of the spindle greatly determine the machining accuracy of the machining center;thus, it is an important factor that affects the overall performance of the machining center. The research is roughly divided into two types, namely, spindle rotation-error measurement based on displacement sensing and that based on optical methods. These measurement methods aim to determine the radial rotation error of the spindle by measuring the displacement change inthe spindle at a certain angle. These two measurementmethods exhibitgood measurement performancewhen the spindle is running at a low speed. However, when the spindle is operating at a high speed, measurement becomes difficult because of the limitation in the sampling frequency. According to the summary of the methods thatmeasure the spindle rotation error, we propose a measurement method for spindle dynamic rotation error based on target trajectory tracking. Finally, we summarize the existing technology and discuss future research prospects.

Hyper-elastic boom-deployable mechanisms are increasingly used to support and collapse the optical film of space satellites.To avoid fatigue crack caused by stress concentration during operation and to prolong service life, the stress in the hyper-elastic boom after flattening under different layup operations was studied. First, a finite-element model of a single lenticular honeycomb(SLH) boom was established, and a non-linear numerical analysis after flattening was performed using the explicit dynamic method. Thereafter, an approximate model of the radial basis function(RBF) was established using the maximum stress after flattening under different lamination modes of the SLH boom. Finally, the Neighborhood Cultivation Genetic Algorithm in ISIGHT software was used to optimize the layup angle. Simulation results show that the error values of the stresses between the finite-element and RBF results are less than 9.95%.According to the resultsobtained,the SLH boom is laminated with two layers, with the optimal layering method constituting the first and second layers, 81.962° and 82.671°, respectively.

To realize the trans-scale and high-precision measurement of microstructures, as well as to perform multiple geometrical parameter characterizations of some high-aspect-ratio structures, a nano coordinate measurement system was developed based on a nano measurement machine and a micro tactile probe. The mechanical, electrical, and software interfaces between the probe and positioning platform were designed. After the integration of the probe and positioning platform, the measurement system was calibrated using a standard micro ball. To ensure traceability of the measurement results, the laser sources of three interferometers in the positioning platform were also calibrated using the laser beat frequency. Finally, ultra-high steps with heights of 10 μm and 2 mm and the sidewall angle of a silicon arm were measured using the developed system. The experiments indicated that the system can accurately measure large-size structures and complex MEMS devices.

In the active optics (AO) system of large closed-looped survey telescopes, wavefront sensing based on defocused star images is necessary. To realize the effective operation of the AO system, this study investigates the relationship between the boundary area of a defocused image and wavefront sensing accuracy. First, a curvature sensing (CS) algorithm is introduced according to the irradiance transport equation of an electromagnetic wave. Then, by considering astigmatism and coma as examples, the influence of typical boundary conditions on wavefront sensing is theoretically investigated. Subsequently, the influence of different pixel resolutions on the completeness of information is analyzed. Finally, under the limit sampling rate, the wavefront reconstruction error is simulated and analyzed. The experimental results show that the error introduced on the edge of a unit circle is approximately 35% (without iteration).When a defocused star point image is below 100 pixels, its ability to carry information rapidly decreases.When it is close to 100 pixels, the detection error of the coma is 0.08 wavelength, which is approximately 14% lower than that at 400 pixels. Therefore, considering the error convergence effect of closed-loop active optics and ignoring the boundary effect can satisfy the requirement of active optical wavefront sensing in the case of a low time-spatial resolution, that is, at a low sampling rate. However, in the case of measurements with a high time-spatial resolution, the influence of the edge cannot be ignored.

In this paper, a piezoelectric piston gas compressor with a flexible drive is proposed to improve the gas driving capacity of a volumetric piezoelectric fluid driver. The operation characteristics of the gas compressor are theoretically and experimentally studied. A vibration model of the flexible support piezoelectric vibrator is developed. The effects of driving frequency and voltage on the vibration characteristics of the piezoelectric vibrator are analyzed through simulations. The flexible support piezoelectric vibrator is combined with a piston pump cavity to construct a piezoelectric piston gas compressor with a flexible drive. A flow-pressure calculation model of the gas compressor is developed to analyze the influences of cavity height and piston amplitude on the compressor output performance. The system resonance principles can improve the gas drive performance of the compressor. In addition, certain structural parameters of the system can optimize the compressor performance. Test results show that the gas output flow and pressure of the compressor are approximately linearly related to the driving voltage and significantly affected by the driving frequency. The best system output performance is obtained when the system operates in the resonance state. When the driving voltage of the piezoelectric vibrator is 220 V, the gas output flow rate and pressure of the compressor are 190 mL/min and 20.3 kPa, respectively.

In this study, a vector propulsion mechanism with the UPR-UPU-UR configuration was investigated to develop an underwater vehicle propulsion device with space attitude adjustment, high precision, and large torque power transmission. First, the degree of freedom of the vector propulsion mechanism was calculated using screw theory, and the position model of the vector propulsion mechanism was constructed using an analytical method. Second, the relationship between the output vector and input vectors of the vector propulsion mechanism was derived, and the particle swarm optimization algorithm was used to calculate the position positive solution of the vector propulsion mechanism. Third, the singularity of the mechanism was evaluated using the Jacobian matrix corresponding to the angle of velocity and the characteristics of the mechanism, and the workspace of the vector propulsion mechanism was calculated. Finally, the experimental platform of the vector propulsion mechanism with the UPR-UPU-UR configuration was built, and experiments were conducted. The results of numerical simulations and experiments demonstrate that the precise position and attitude of the mechanism could be obtained when the absolute error was less than 0.001°, and the maximum error between the experimental and theoretical values of the vector propulsive performance was 5%, which proved that the mechanism had good motion characteristics and propulsive performance. The vector propulsion mechanism adjusted the thrust direction of the propeller by varying the length of two branch chains, such that the robot could exhibit various motion modes, such as yaw, pitch, and roll, which helped in achieving the complex underwater motion.

In this study, the measurement accuracy of an embedded time-grating sensor was characterized for different mounting modes. Based on the sensing mechanism of the embedded time-grating sensor, the effects of variations in the mounting air gap, radial mounting, and axial mounting on the measurement accuracy of the sensor were analyzed, and the sensing measurement model was established for different mounting modes. Simulation and experimental verifications of the mounting modes of the sensor were performed. The simulation and experimental results showed that different mounting modes influenced the measurement accuracy of the sensor, with radial mounting having the most significant effect on the measurement error of the sensor, followed by air gap spacing and axial mounting. This study has significant theoretical and practical implications in guiding the installation, minimizing the installation errors, and improving the measurement accuracy of embedded time-grating sensors.

A 3D point cloud model that takes curvature and variation of the normal vector as metrics is not dual with regard to a certain reference plane. This can lead to the production of numerous meaningless topological features from the point cloud Morse-Smale (MS) complex extracted based on Morse theory, severely restricting the recognition efficiency of model features. To address this problem, the concept of a single complex topology model is proposed to avoid the extraction of meaningless features. Based on the characteristic line importance measurement method of the MS complex and the homomorphic shrinkage algorithm, the characteristic line importance measurement method and topology simplification algorithm of a single complex topology model are derived. Furthermore, the model transition features are difficult to retain in the process of topology simplification; to address this, by setting thresholds based on the single complex construction and persistence simplification theory, the saddle points that lead to the generation of critical lines that cross the contours or are off the contours are filtered and deleted. The protection of transition features is achieved in the simplification process. The algorithm was experimentally validated on several typical 3D point cloud models. The results and analysis show that, in contrast to existing topological feature extraction methods, the extraction and simplification algorithm of the single complex topology model successfully avoids extracting several meaningless features. The time efficiency and data compression rate respectively increase by 52.22% and 5% or more. With this blend feature protection method, the identification rate of blend features reaches 100%. A large number of experimental data and a series of subsequent analyses demonstrate that this method significantly improves the feature recognition efficiency of the 3D point cloud model. Moreover, it effectively alleviates the problem of incompleteness and fracture in transition feature lines in conventional algorithms.

Cross-domain face recognition (FR) has always been a research hotspot in the field of face recognition. It has high application value and development potential in real applications such as security and criminal investigation. The existing cross-domain face recognition methods usually establish the correlation between different domain faces in the image space or latent subspace, but ignore the intrinsic relation between the two, which easily leads to the loss of inter-modal correlation information. In order to solve this problem, in this paper, we propose a novel method, called Cross-Domain Representation Alignment (CDRA). CDRA algorithm explores the correlation between different domain face data in the face image space and latent space. First, in order to reduce information loss, the CDRA algorithm can learn the latent feature representation containing discriminant information by reconstructing the face in a single domain. Then, in image space, CDRA algorithm is used to cross domain from different domain latent features. In the latent space, CDRA directly aligns the latent feature representations of different domain by aligning the latent Gaussian distribution of different domain data, which promotes the feature representation to learn the cross domain information of different domain faces in different spatial dimensions and levels. Experimental results indicate the average face recognition accuracy rate of CDRA is 97.2% on Multi-Pie dataset, and 99.4% ± 0.2% on CASIA NIR-VIS 2.0 dataset. Simultaneously, the efficient cross-domain face synthesis is realized. The learned latent features of our CDRA method can obtain the essential cross-domain information in both image space and latent subspace for cross-domain FR task, which can effectively improve the cross-domain face recognition.

A multi-target positioning system based on a digital elevation model is proposed to realize the accurate positioning of multiple targets using the on-board optoelectronic equipment of an UAV. A multi-target positioning model based on the target vector is developed. The pixel coordinates of each target in the field of view are obtained through target detection. The boresight vector of each target can be obtained. The measurement data of each sensor in the drone and optoelectronic reconnaissance platform are combined to calculate the geographical position information of each target. For a flight height of 3000 m, the experimental results show that the positioning error for the main target is approximately 16 m, whereas that for the secondary target is approximately 26 m. The effect of the improved filtering model is analyzed. After filtering, the positioning error of the main target is reduced to 7 m, whereas that of the secondary target is reduced to approximately 11 m. The verification of the flight experiment demonstrates that the engineering effect is basically consistent with the simulation experiment result. This method has the advantages of good real-time positioning, a small error, and facile engineering application.

Soil temperature is an important variable in Earth sciences. The temporal and spatial variations in soil temperature are affected by numerous factors, resulting in various challenges in soil temperature prediction. For soil temperature prediction, the data-driven machine learning method is valuable and can be an important complement to physics-based process models. However, no extensive studies have been carried out on the importance of environmental factors on soil temperature. In this study, a data-driven XGBoost-LSTM method is proposed. The weights of the meteorological inputs are computed based on XGBoost, and then, the combination of meteorological inputs based on their weights is applied to obtain an optimal model by the LSTM method. An experiment is carried out at two stations in China (Changbai Mountain and Haibei). The most accurate performance for soil temperature estimation is attained, with highest values of NS = 0.932, WI = 0.983, and LMI = 0.729 and lowest values of RMSE and MAE of 2.234 and 1.716, respectively. These results show that the proposed model is generally superior to other state-of-the-art predictive models.

To address the limitations of the scale invariant feature transform (SIFT) algorithm and reduce the computational burden of the Ane-SIFT (ASIFT) algorithm in scenes with large affine deformations, a fast image matching algorithm based on approximate-ane-SIFT (Fast-AASIFT) is proposed. Fast-AASIFT has a clearer physical meaning than the ASIFT algorithm. First, Fast-AASIFT recovers original images as rectified images by performing inverse affine transformations. Then, it performs feature point extraction and SIFT description on the rectified images. Finally, it performs SIFT optimization matching. The experimental results demonstrate that, in scenes with a large affine deformation, Fast-AASIFT can still match enough feature points, with a peak matching error of <2.5 pixels and an average matching error of <1.2 pixels. This proves that the anti-affine deformation ability of Fast-AASIFT is equivalent to that of the ASIFT algorithm, which is significantly better than that of the SIFT algorithm. Furthermore, the time consumed by Fast-AASIFT less than 30% of that consumed by the ASIFT algorithm; thus, it effectively addresses the time-consumption problem of the ASIFT algorithm. Obviously, Fast-AASIFT not only maintains good robustness against affine deformations but also greatly improves computational efficiency; consequently, it is of great value for applications such as scene reconstruction and recognition.

Ships are vital marine targets. Real-time ship target detection based on aerial remote-sensing images is useful in civil engineering as well as national defense security. Based on the engineering application background, land-sea segmentation and ship detection was investigated in this study using aerial remote-sensing images. Gray-scale information was used to segment the remote-sensing images with adaptive threshold; the morphological operator and hole-filling technique were combined to achieve land-sea segmentation. Furthermore, ship geometry features were used to detect the straight-line segment, and K-means density clustering was applied to complete the ship target detection. The experimental results showed that for land-sea segmentation, the land detection rate was 95.8%, the land detection error rate was 5.7%, and the land detection accuracy rate was 94.4%. For ship target detection, the detection accuracy rate was 94.1%, and the detection false alarm rate was 3.9%. The proposed technique effectively segmented the ocean and land areas of the images, and the segmentation effect was ideal. The proposed technique had the following advantages: rapid ship target detection, high accuracy, low false alarm rate, simple calculation, and easy engineering applicability.

To address the problems of fuzzy entropy-based multilevel threshold segmentation methods, such as insufficient fuzzy characteristics, high computational complexity, and poor automaticity, a multilevel threshold segmentation method for high-resolution panchromatic remote sensing imagery is proposed based on interval type-2 fuzzy entropy. First, a ridge-type fuzzy membership function is applied to construct an interval type-2 fuzzy set, and interval type-2 fuzzy entropy is defined in the multilevel image segmentation scene based on the constructed fuzzy set and the number of thresholds. Then, qubits encode a fuzzy parameter set as quantum chromosomes, and several quantum chromosomes are set to form the initial population. In addition, the defined interval type-2 fuzzy entropy is adopted as the fitness evaluation function to evaluate the fitness of individuals in the population, retaining and recording the best individuals. In the proposed evolutionary strategy, the dynamic rotation angle mechanism of quantum rotation gates is applied such that the population can automatically determine the optimal combination of fuzzy parameters with better adaptability and efficiency. Based on this, the multilevel threshold is obtained by the principle of maximum fuzziness, and the optimal multilevel threshold segmentation of the image is realized. In an experiment, a multilevel threshold segmentation method based on maximum entropy and fuzzy entropy was employed as the comparison algorithm to segment high-resolution panchromatic remote sensing images with different ground objects. The averages of the experimental evaluation results show that the proposed method can obtain better segmentation results while reducing the computation time. The area weighted variance is reduced by 39.7%, the Jeffries-Matusita distance is reduced by 14.7%, and the running time is 6.403 s. The method can meet the requirements of high-resolution panchromatic remote sensing image segmentation for spatial continuity and spectral uniformity, resulting in high real-time performance.

The sparse recovery algorithm needs to perform grid quantization processing in the angle space when DOA estimation is performed. Aiming at the problem that the quantization error introduced by the quantization process affects the estimation performance, this paper introduced the quantization error into the second-order moment model of the array output through the first-order Taylor expansion of the steering vector. Based on this model, an OMP algorithm that used noise subspace vectors to modify was designed to jointly estimate DOA and quantization error. The new algorithm based on the array covariance matrix was slightly sensitive to the dependency on the number of snapshots, but it made up for the lack of DOA resolution of the greedy algorithm and did not require the number of sources to be predicted. At the same time, the amount of calculation was compared with the existing convex based on the Lp norm constraint. The optimization method was greatly reduced. Simulation experiments verify the effectiveness of the proposed algorithm.