This PDF file contains the front matter associated with SPIE Proceedings Volume 12166, including the Title Page, Copyright information, and Table of Contents.

Ta₂O₅ thin films are widely used in optical and microelectronic industry because of its superior optical and mechanical properties. In this paper, single-layer Ta₂O₅ thin films were prepared by APS plasma assisted electron beam evaporation deposition. Based on the theory of ion energy transfer, the selection criteria of APS process parameters were established. By optimizing APS source parameters, Ta₂O₅ thin films with different characteristics were prepared. The spectral and refractive index dispersion of Ta₂O₅ thin films were analyzed by Cary7000 spectrophotometer. The stress and surface roughness of Ta₂O₅ films were analyzed by Zygo interferometer. Experiment and analysis results showed that the characteristics of Ta₂O₅ thin films are closely related to APS plasma assisted processing parameters. The discharge current and bias voltage of APS source have great influence on the stress and surface roughness of Ta₂O₅ thin films, but have little influence on the spectral characteristics and refractive index dispersion. The influences of preparation parameters on the properties of Ta₂O₅ thin films were analyzed and optimization fabrication parameters were obtained.

Ion beam figuring has the advantages of non-contact, high certainty, and removal function’s long-term stability. It is the primary processing method of ultra-high precision optical elements such as lithography objective lens. In general, the IBF process does not require high motion precision of the moving mechanism. However, when the shape of the optical components changes from a plane to a curved surface and the RMS value of the surface accuracy is better than 0.3nm, the influence of the motion accuracy cannot be ignored. The paper focuses on the influence mechanism of IBF equipment’s motion accuracy on the shape of removal function, removal rate, and other essential parameters under the three-axis figuring mode of curved surface elements. We have discussed the sub-nanometer precision generation requirements of curved elements for removal function positioning accuracy and motion accuracy of polishing equipment. The feasibility of ion beam triaxial machining for the sub-nanometer precision generation of curved optical elements is demonstrated by analysis and calculation. The results show that under the condition of ensuring the micron motion accuracy, the three-axis ion beam figuring equipment can also process the curved surface’s surface to the ultra-high precision that the RMS value is better than 0.3nm. Still, it is difficult to achieve the machining goal only by improving the motion precision of the machine tool’s motion axis, so it is necessary to control the components’ initial profile error.

Due to Metal mirror’s low hardness, we coat a layer of nickel-phosphorus (NiP) layer on the surface and then polish NiP layer simpler through various methods. However, when using CeO₂ slurry in Magnetorheological Finishing(MRF), the NiP layer will appear "wise-tail" phenomenon. Based on the "double-edged circle" model, we analyze the shear force of the contact surface and optimize the polishing abrasive particles such as SiO₂. In this way, we could obtain a surface without "wise-tail" and the surface roughness reaches 1.897 nm in Ra. By orthogonal experiments, we optimize the process parametersts and certify the stability and linearity of the removal function.The experiment results prove that the MRF fluid compounded with SiO₂ could satisfy the requirements of NiP layer in magnetorheological machining.

Starting an experiment about image stitching of blade-shape hole. Firstly, Focusing on algorithms of feature point detection. Harris, SIFT, and SURF are the main feature point detection algorithms. All of them are used to detect feature points on the same image, and analyze the precision of it and time-consuming. Then compare algorithm of the points match (brute force and KNN, brute force and crosscheck, K-D tree and KNN) by SSIM and line profile. Image fusion is determined based on pixel weighting, and transformation matrix is gained by RANSAC. Last, the system uses the Harris algorithm to detect feature points, and uses brute force and the KNN algorithm to match them. Verify the algorithm of blade-shaped hole image stitching is effective, and lay foundation for the next work of this project.

Vehicle-mounted telescopes will suffer a lot of outside interference during the transportation and use. Such as the vibration and temperature difference. Aiming at the problems, this paper optimized the design of the supporting system for the main mirror of the 800mm telescope. Firstly, the position of the main mirror support point and the size of the support force are optimized and analyzed by using the ANSYS. In order to analyze the vibration during transportation, the vehicle transportation process is simplified to a random vibration model, and analyze its mode of vibration. At the same time, the influence of temperature difference on the accuracy of the main mirror surface during use is analyzed. The results show that after optimization of the main mirror support point and the support force, the RMS (Root Mean Square) of the surface accuracy under calculation is less than 1/30λ(λ=632.8mm), the modal frequency of the main mirror(207Hz) is greater than the general vehicle vibration frequency(20Hz), it can effectively avoid resonance. On the basis, loading random acceleration excitation, the maximum stress is only 49.8Mpa, which is much smaller than the yield strength of the material used(>355Mpa). It can guarantee primary mirror accuracy well. And under the temperature difference of 5K, the maximum deformation of the main mirror is 0.106μm, which can well satisfy the design requirements.

In order to solve the problem that the extra removal layer and the motion characteristics of the machine tool are difficult to meet the processing requirements of ion beam figuring, an ion beam figuring method based on new controllable ion source is proposed. By changing the working parameters of the ion optical system, the timing and duration of ion beam extraction are controlled in real time. The influence law of the machine tool motion acceleration in the process is analyzed theoretically, and then a new ion beam figuring method is proposed for the lack of dynamic performance. By adjusting the working parameters of the ion source developed by ourselves, the pulse duty ratio is continuously adjustable from 0 to 100% , and the pulse frequency is continuously adjustable from 1 to 1000 Hz. The sample is Φ100 mm monocrystalline silicon plane mirror. Firstly, the long-time stability of the new ion source was verified by line- scanning experiments, and then the error of 14.5 mm wavelength was etched with the axis of motion at a constant speed. The results show that the technology can make up for the lack of motion acceleration and avoid the extra removal layer, and have a wide range of potential applications in high precision quality adjustment, special surface treatment and so on. It is expected to promote the progress of ultra-precision machining technology.

Elastic emission polishing (EEM) has excellent polishing surface quality, and shear thickening can improve polishing removal efficiency. This paper mainly studies the effect of shear thickening slurry on EEM. Firstly, the polishing device and removal mechanism based on shear thickening are introduced. Then a single-point contrast polishing experiment was carried out. It is found that the shear thickening slurry can improve the removal efficiency of ordinary EEM, the former is about 2.7 times that of the latter.

To construct freeform surface that best fits the normal field, an efficient numerical method is put forward in this paper. In this method, the normal vectors are constructed by two adjacent orthogonal tangent vectors at each point, and then the normal vectors are substituted into Snell's law to obtain the nonlinear equations describing the surface coordinates. Finally, the continuous and accurate freeform surface can be obtained by solving these nonlinear equations. The simulation results show that the proposed method provides lower RSD and normal deviation than the traditional freeform construction method.

Digital image correlation (DIC) is a widely used non-contact measurement method for the deformation of a diffused object surfaces in the experimental solid mechanics. In specular surface figure measurement, the rapidly developing deflectometry using sinusoidal fringes can achieve a sub-micron level accuracy. The fringe pattern used in the phase shift algorithm is sensitive to noise, but the speckle image used in the DIC technique is robust against noise. In this paper, three kinds of speckle deflectometry methods, which are speckle pattern defletometry(SPD), speckle pattern shifting deflectometry(SPSD) and dynamic speckle deflectometry(DSD), are introduced to measure the figure of specular surface. The principles of the three methods are introduced in detail, and the experimental results are given in order to illustrate the measurement accuracy.

Phase Measuring Deflectometry (PMD) with high dynamic range and high robustness is a more powerful method to achieve the integration of optical components fabrication and testing, which would avoid the secondary positioning error caused by the removal of optical components in off-line status. However, for surface shape measurement of transparent element, the captured fringe patterns are superposition of the patterns reflected from the front and rear surfaces, leading to the failure of the traditional phase-shifting algorithm to extract the true phase distribution. In order to separate the superposed fringe patterns, the limitations of the existing methods such as UV deflectometry and polarized light deflectometry, which require special light source and polarizer filter, are expensive instrument and complex processes. Moreover, only the front surface shape can be reconstructed at a time, which lowers the efficiency of the measurement. To achieve in-situ measurement of the front and rear surface shapes of the transparent element simultaneously, the combination of PMD and power spectrum estimation is used to separate parasitic fringe patterns. And the front and rear surface shapes are reconstructed using ray tracing and non-linear optimization. The feasibility of the proposed method is demonstrated by numerical simulation. In the experiment, the transparent element with a thickness of about 10 mm is detected and the front and rear surface shapes are reconstructed. The front and rear surface shapes differences of the transparent element with a diameter of 77.8 mm are 173 nm in RMS and 212 nm in RMS, respectively.

According to the principle of differential geometry, the surface power distribution of a continuous surface can be solved analytically. However, the reverse engineering from surface power distribution to surface geometry relies on numerical optimization methods. This paper proposes an efficient and high-precision surface construction method for rotationally symmetric multifocal optics design. Based on the requirements for the surface power distribution in the radial direction, this method calculates the relative offset of the curvature center coordinate along the x-axis direction for each point on the cross-section curve, then calculates the first-order derivative and the second-order derivative of the cross-section curve and the corresponding curvature radius and surface power. The deviation of the designed surface power from the target power is used as a measure of the accuracy of the multifocus optics design. This method is suitable for any rotationally symmetrical multifocal distribution pattern, including Gaussian multifocal distribution and step multifocal distribution, and has the potential to extend to the free-form multifocal surface design.

At present, most polarization detectors are based on polarization filtering methods. The principle is that different detection areas only transmit a certain component of polarized light. Stokes parameter reconstruction is achieved by obtaining four or six polarized components of incident light. However, the sensor of this polarization detector can only receive 50% of the incident light energy. To improve the energy utilization of the polarization detector. Based on the flexible modulation capability of the dielectric metasurfaces to phase. We designed a metasurface that can separate incident light by polarization, and then focus the polarization components in different ways. The light energy received by the sensor is only limited by the transmittance of the metasurfaces. The polarization detection metasurface is composed of three metalenses. The numerical simulation result is that the error of Stokes reconstruction parameters is less than 0.03, and the transmission efficiency is higher than 60%. This detection metasurface improves the energy efficiency of traditional polarization detectors. This kind of polarization detection metasurface has great significance to the development of ultra-miniaturization polarization detection systems.

The fabrication of high-precision mold core can make the optical elements of freeform surface be mass-produced by plastic injection or glass mold pressing. However, diamond tools cannot directly cut ferrous materials due to rapid wear of tool. Therefore, the technology of ultrasonic assisted diamond turning is proposed to process the freeform surface on the die steel in this paper. The characteristics of vibration servo machining are analyzed and discussed by theoretical modeling and digital simulating. After fabricating freeform surface on die steel by this technology, the value Ra of surface roughness can be achieve to 5.905nm.

Introducing adaptive optics such as spatial light modulators (SLMs) can help troubleshoot the ‘one-to-one’ test mode of interferometric null surface figure tests for aspheric and free-form surfaces. However, the phase controlling accuracy of methods based on birefringence effects and phase-wrapping is too low for applying SLMs in interferometric surface figure tests. To this end, the flexible phase control method based on computer holography for SLMs is presented. The difference of phase control principle between the computer holography-based method and the birefringence effects and phase-wrapping based method is discussed. Phase control accuracy of the proposed method is about λ/30 (λ=632.8 nm) root-mean-square (RMS) according to the theoretical analysis result. The inherent wavefront error of a SLM Holoeye^TM LC 2012 is significantly nulled to from RMS 0.426λ to RMS 0.015λ, which verifies the high phase controlling accuracy of the proposed method. By troubleshooting the phase control accuracy issue, the presented method is promising for promoting the application of SLMs in interferometric surface figure tests.

In order to improve the fatigue evaluation system of the satellite in the transportation environment, a fatigue analysis method of satellite transportation environment based on the digital twin was explored in this paper. At first, the basic theories and method of fatigue analysis are introduced in detail. The fatigue damages of the satellite in the transportation environment are evaluated by using the random vibration and sinusoidal vibration stress fatigue analysis separately. And then, digital twin fatigue models based on the acceleration response of random vibration and sinusoidal vibration are established. Finally, fatigue models are utilized to analyze the fatigue damage to the satellite in the transportation environment and are compared with the stress fatigue models before. The results show that all the fatigue damage keeps in accordance with each other verifying the accuracy of the fatigue model based on the digital twin presented in this paper.

Taking the grinding process of curved optical parts as the research object, a neural network based the adaptive control method is proposed. MATALAB is used to design the controller and simulate the grinding process. The results show that the method can effectively solve the fluctuation of the grinding force in the grinding process, and the controller has good dynamic characteristics, The purpose of the optimal material cutting rate in the grinding process is realized, and the surface processing quality of optical parts is improved.

Complex electromagnetic environment in the future battlefield requires spectrum sensing equipment to have broadband and high-resolution measurement capabilities. This paper proposes a microwave photonic frequency measurement method based on optical spectrum operation and stimulated Brillouin scattering. The use of optical spectrum operation is to realize generating pump light in a large range, so as to realize the excitation of stimulated Brillouin scattering in a wide spectrum range, and further the stimulated Brillouin scattering is employed to realize high-resolution frequency sensing. The principle of this method is given and the feasibility of the method is verified experimentally. Experiment results show the measurement capability of the proposed method covers a frequency range of 0.03–40 GHz with a resolution of 25 MHz. The proposed method can effectively support high-resolution frequency sensing in complex electromagnetic environments.

In this paper, a new deformation measurement method based on fiber bragg grating strain monitoring is proposed. Aiming the four-sided fixed elastic thin plate structure, an algorithm for reconstructing deformation field based on fiber bragg grating strain monitoring is proposed. The mechanical model of elastic thin plate was established. According to the degree of deformation, it can be divided into small deflection deformation and large deflection deformation. The relationship between surface strain and curvature of a thin plate was deduced. On this basis, a curvature iteration method was proposed to reconstruct the deformation field of thin plates. Aiming at the non-linear problem of large deflection deformation, a subsidence correction algorithm based on curvature iteration method was proposed. The experimental platform for strain monitoring of thin plate and several different loading conditions was established. Based on the discrete point strain data obtained by fiber bragg grating strain monitoring, the strain field is reconstructed to obtain detailed strain information. Based on the curvature iteration method and strain information, the deformation curves of thin plates under different loads are obtained. The experimental results show that the average error of reconstruction of the elastic thin plate structure is 6%, and the reconstructed results can accurately reflect the deformation surface of elastic thin plate. It also provides a basis for the study of deformation field reconstruction under complex constraints.

This paper presents a methodology for optimization of the scanning path of iterative adaptive grid generation algorithms in metrological scanning probe microscopes (SPM). This method equips the traditional non-equidistant (adaptive) grid point generation algorithm for the measurement of large smooth optical workpieces with a space-filling curve solution of the traveling salesman problem. We will show that the combined algorithm can reduce the total measurement time efficiently and enables faster surface scanning speed for large industrial workpieces while preserving adequate surface information for performance evaluation after surface reconstruction comparing to contemporary paths. The path generation algorithm also promises necessary information locality and better surface reconstruction completeness. The algorithm is not subject to instrument restrictions and can be implemented in any coordinate measurement systems with xy scanners. Experimental verifications are conducted on a typical parametric surface to show the efficiency of the proposed algorithm.

At present, aspheric mirrors are being used more and more widely in various optical systems, and the quality and precision requirements of the components are gradually increasing. Aspheric mirror processing process is divided into milling and grinding molding, aspheric grinding, finishing treatment, coating, etc. Polishing is an important process to improve the precision of the components, and various polishing methods are rapidly developing, mainly ion beam processing, magnetorheological processing, magneto-jet polishing and airbag polishing. Among them, airbag polishing belongs to the flexible polishing, his advantages are uniform and controllable pressure, adjustable polishing speed, can adapt to more complex surface characteristics and high polishing accuracy, in the finishing application prospects. But the airbag polishing path planning has been a problem of choice, common path planning methods and raster path, concentric path, spiral line path, sine spiral line path, random path and so on. The first three paths are too regular because of the trajectory, so they will introduce high and medium frequency errors in processing. The sinusoidal spiral path eliminates some of the medium and high frequency errors by introducing the sinusoidal oscillation of the airbag on the basis of the spiral path. The random path, on the other hand, avoids the generation of the mid-high frequency error by avoiding the appearance of a regular route.

This paper proposed a tool path (TP) method for machining off-axial multi-mirrors optical systems considering machining accuracy. The calculation of residual error and chord error were modeled. The minimum number of cutting contact points (CCPs) in circumferential direction and the minimum feed for TP can be obtained using the calculation model of chord error and residual error. The curvature is an essential factor for modeling of residual error and chord error. However, the surface equations of the mirrors are complex. It is difficult to calculate the curvature of the surface by derivation. In this paper, the Frenet frame was used to approximate the curvature. A simulation of TP generation for an off-axial four-mirror anastigmat system using the proposed method was conducted. The simulation results of residual errors and chord errors of the four mirrors showed that the proposed TP generation method is effective.

A subwavelength interference nanolithography scheme is numerically demonstrated based on a metallic-insulator-metallic waveguide (MIMW) combined with two symmetrical surface plasmon unidirectional couplers. The unidirectional coupler is composed of a conventional nanoslit with a nanochannel excavated form onside of this nanoslit. Simulation results show that a feature size of 50 nm half-pitch and a depth of 100 nm uniformity pattern can be obtained at the working wavelength of 365 nm with the proposed scheme. This proposed scheme can enhance the intensity of the interference pattern with higher contrast and larger field depth which is advantageous to the fabrication process in practical applications.

Precision glass molding (PGM) technology provides a possibility of rapidly manufacturing optical lenses with large volume, high precision, and complex shapes. It is difficult to fabricate micrometer-level structures in common PGM mold materials like tungsten carbide (WC) using a grinding/polishing approach. But no literature has pointed out how to design the dimensional tolerance of molds to ensure the molding accuracy in the case of the molds working in a temperature range of several hundreds of degrees Celsius. In this paper, a theoretical model is established based on the general molds designed sizes. The finite element method (FEM) and the PGM experiments were carried out. The measuring results of the decenter and tilt of the molded lens indicated that the optimization was effective and could be applied to other PGM experiments.

This paper proposes a channelized receiver based on microwave photonic, the system generated two coherent optical frequency comb with different free spectral ranges. By adjusting carrier frequency, a different frequency interval is formed between the local frequency comb and the optical comb modulating the signal to be measured. After the two optical combs pass through the optical demultiplexer respectively, the corresponding channel is beat frequency and the heterodyne detection is realized by converting the signal to intermediate frequency.The scheme proposed in this paper realizes the measurement frequency range of 2GHz~18GHz, has the instantaneous bandwidth of more than 7GHz, and the average error is less than 50MHz

During the process of ultra-precision diamond fly-cutting machine tool, due to internal heat sources, the thermal deformation of various parts of the machine tool has severely restricted the accuracy of the machine tool. Therefore, in order to improve the accuracy of the machine, it is necessary to study the mechanism of ultra-precision diamond flycutting machine tool. First, an intelligent monitoring system is arranged for the fly-cutting machine tool to collect temperature signals. Then the heat generations at the spindle motor and linear motor are calculated, and the thermal boundary conditions of the machine tool are analyzed. Finally, the finite element model of the flying-cutting machine tool was established, and the thermal-structural coupling analysis and calculation were carried out.The results show that: under the action of the internal heat source, the maximum deformation of the machine tool in the cutting direction is 6.8μm, which occurs at the spindle motor. The maximum deformation of the tool tip is 1.02μm in the cutting direction. Eventually, the machine thermal error under the tool tip and the work bench is 1.42μm. This research provides a basis for the research of machine tool thermal compensation and the improvement of the accuracy of machine tool.

In order to obtain the microstress bonding curing process and its properties of precision optical components, various curing process schemes of XM-31 adhesive were designed and tested. Firstly, a mirror group simulation structure and hardness testing tool were developed to study the effect of the amount of six vulcanizing agents on the solidification rate of XM-31 adhesive. Secondly, the XM-31 adhesive with the best amount of curing agent was solidified under different temperature conditions.Densitometer, stretch machine, hardness tester and microscope were used to test the material properties of XM-31 adhesive after curing, so as to obtain the bonding performance of XM-31 adhesive under different solidification temperature conditions.Finally, the optimized process is applied to the bonding of optical elements in a star - sensitive lens. The results show that the optimum ratio of matrix adhesive and vulcanizing agent of XM-31 adhesive is 100:3 ~ 100:4. The volume shrinkage rate of XM-31 adhesive under four curing temperatures were 1.82%, 1.75%, 1.63% and 1.43%, respectively. The curing temperature will significantly improve the adhesive process efficiency, but the adhesive strength and hardness will be reduced. The surface shape index PV and RMS value of a certain type of star sensitive lens can meet the design requirements, when using the preferred process of XM-31 adhesive assembly.

Cangzhou-Hengshui-Nangong subsidence, located in the southeast of North China Plain, has become the largest subsidence bowl with the maximum rates of subsidence and influenced regions in North China Plain. We employ PSInSAR and SBAS algorithms to retrieve deformation time series of Cangzhou-Hengshui-Nangong subsidence area, in order to research its spatial-temporal distribution characteristic of subsidence. Meanwhile, we discuss the correlation between groundwater depth, groundwater funnel areas and land subsidence and the deep-confined groundwater overexploitation bring about large-scale subsidence.

In this paper, we address the problem of direction finding using coprime array, which is one of the most preferred sparse array configurations in the radar community. Motivated by the fact that conventional algorithms based on the convex relaxation is computationally expensive, we propose a non-convex accelerated structured alternating projection-based direction-of-arrival (DoA) estimation approach without the need to solve a semi-definite programming. The low rank Hankel matrices constructed by the sub-arrays of the virtual uniform linear array (ULA) and the sparse hole signal respectively correspond to the sets of low-rank matrices and sparse manifolds. After initialization, we update the estimations by iterative projections between the two sets, which employs an accelerated low-rank approximation method. Once the desired Hankel matrix with DoA information is recovered, simple subspace-based spectral estimation algorithms can be applied to obtain the target directions. The estimation performance in terms of root-mean-square error is examined and the superiority of the proposed method over the competitive approaches in the computational cost sense is also demonstrated.

Due to the multi-antenna structure in both transmitting and receiving arrays, the imaging performance of a Multiple-Input Multiple-Output (MIMO) imaging radar is quite sensitive to the inter-channel errors. To compensate these errors, an iterative calibration method for the 3D MIMO imaging radar is proposed based on the minimum entropy technology in this paper. This method can achieve a joint estimation of the amplitude and phase errors and be performed based on the echo of the natural scene instead of placing prominent point targets. Finally, the correctness and validity of the proposed method are verified by a simulation and an experiment.

Hand gesture recognition has long been a study topic in the field of Human Computer Interaction. Compared with traditional camera-based recognition systems, radar can realize dynamic hand gesture recognition at long distances and in low light conditions by exploiting the micro-Doppler (m-D) effect. However, for some static and complex hand gestures, the m-D-based recognition methods are rendered impotent. In this paper, we present a method based on 2-D synthetic aperture radar (SAR) imaging to distinguish nine kinds of static hand gestures representing the numbers 1-9. A cost-effective 77 GHz mm-wave radar is used to achieve imaging and data acquisition of different gestures. Finally, two classifiers including classic machine learning and deep learning are used to evaluate the effectiveness of the proposed method. Experimental results demonstrate that the proposed method can guarantee a promising recognition accuracy.

In order to effectively overcome the limitations of traditional gesture recognition technology, a method of gesture recognition using millimeter wave radar is proposed. First, according to the introduction of the millimeter-wave radar system and the description of the echo model, the millimeter-wave radar is used to collect the measured data; then the average cancellation method is used to suppress the clutter of the measured data, and the joint time-frequency analysis technology is used for effective feature extraction; The extracted features are used as the model input, and a lightweight convolutional neural network model is improved. Its recognition rate is over 96%, and it has good recognition performance.

In the real-time image processing system of the airborne infrared camera, how to process a large amount of data and information in a limited time and effectively meet the real-time requirements of the system is a problem that needs to be solved as soon as possible. Based on this, this article meets the needs of system modularization design ideas, and proposes a real-time image processing system design based on CMOS, using DSP and FPGA as the core devices to realize the corresponding hardware circuit design, and using high-demand tracking in the real-time image processing system algorithm. The experimental results show that the functions and performance of the designed real-time image processing system can meet the expected demand, and it is practical and reliable.

The sudden interference of highlighted objects in the field of view of infrared imagers causes the overall gray distribution of the image to shift to the highlighted region, leading to the decline of target recognition effect. In this paper, an infrared image processing algorithm which runs by FPGA is introduced. Through histogram statistics and gray value conversion, the infrared image is compressed and the image details are enhanced. Through the recognition and processing of the high gray value region in the histogram, the interference of highlighting can be suppressed, and a good effect has been achieved.

The complexity of the mountainous terrain leads to the large phase gradients of the interferogram, especially in the area with the dense vegetation. And the resulting low coherence brings difficulty to phase unwrapping. To overcome the difficulty of phase unwrapping in these areas and the weakness of a single method, this paper proposes a novel multi-level phase unwrapping method combining the branch-cut and minimum cost flow algorithm. The proposed method is inspired by the principle of “control first then fragments” in surveying, specifically, the unwrapping task is progressively performed from the high-coherence regions to the low-coherence regions by taking the coherence as an indicator. Firstly, the unwrapped phase result obtained from the minimum cost flow method in the high-coherence region is seen as the “control” part, i.e. the initial phases. Secondly, the “fragments” phases are computed using the branch-cut algorithm. The process is repeated several times until the interferom of the whole image is done, and each time a part of the remaining low-coherence region is determined by setting an appropriate coherence threshold then unwrapped based on the previous unwrapping phases. The experimental result using the TanDEM-X Bistatic mode demonstrates the new method has higher efficiency and more reliability than the branch-cut method and the minimum cost flow method.

Target location is a crucial property for the civil and military applications and should be performed with high precision immediately. The signal-level cooperative active location of moving targets is studied with the distributed networked radar (DNR) system employing widely dispersed transmit and receive antennas. By setting up simulation and test scenarios, both a centralized processing and a simple distributed processing form the DNR system are developed and studied, and the estimated target track plots of the simulated and experiment data acquired by the first long baseline DNR system in China is contrasted with that of the ADS-B. Furthermore, a detailed verification analysis for the location accuracy obtained by the DNR system is compared with that of a single radar. The results indict that the high accuracy can be obtained by the proposed method owing to the gains from the having widely dispersed antennas of the DNR system and the signal-level centralized detection. The DNR system achieves a spatial diversity gain to overcome fades in target RCS, and simultaneously, it improves the target location accuracy using coherent and non-coherent processing under the condition of low signal-to-noise ratio when comparing to the individual radar composing the network.

Infrared imaging system compared with visible light imaging system, infrared imaging system can operate in all-weather conditions with high anti-interference ability and the ability to penetrate smoke and haze. The current infrared technology still has the disadvantage of low signal-to-noise ratio and lack high frequency detail information. The most direct way to improve the infrared imaging system is to improve the hardware design by increasing the size of the photoreceptor and the size of the image source, but the manufacturing process is complicated, the cost is high, and the improvement effect is limited. With the development of the field of computer vision, deep learning has become the most effective solution for super-resolution reconstruction[6]. Therefore, in this paper, a Video frame Infrared image super-resolution reconstruction method IVSR (Infrared Video super-resolution) is designed based on the deep learning method and the dual-path operation mode combining the single-frame and multi-frame super-resolution reconstruction algorithms. The innovation of IVSR net lies in that after the optical flow module extracts the inter-frame information of each frame and the current frame, the sub-pixel convolution layer of the multi-frame image fusion module can effectively utilize the sub-pixel information. IVSR can be regarded as two reconstruction processes, the first stage is the output of high-resolution information from the optical flow motion estimation module to the projection module, and the second stage integrates the high-resolution information of each projection module for fusion reconstruction. This method can effectively improve the visual effect of IVSR reconstruction, effectively solve the problem of poor reconstruction quality due to the lack of high frequency detail information. Compared with traditional reconstruction algorithms and other typical deep learning algorithms, reconstructed images of IVSR are more exquisite, with more prominent high-frequency details, no image distortion and significant advantages in objective evaluation indicators.

Synthetic aperture radar (SAR) tomography is a technique that offers new perspectives in the 4-D reconstruction (3-D spatial information plus deformation information), particularly, of urban areas. SAR tomography allows resolving layovers by handle the situation where multiple scatterers exist within a resolution cell. The detection of scatterers is however a critical and challenging issue. Single scatterers can be simply detected by the amplitude dispersion method, but it does not meet the application needs of urban buildings. In this paper, we study a variety of methods for scatterer detection in differential SAR Tomography. We mainly investigate three detection approaches based on signal estimation and detection: sequential generalized likelihood ratio test (SGLRT), sup-GLRT, and fast-sup-GLRT. Firstly, theoretical performances of each method are analyzed in details. Secondly, experimental validation is carried out through simulation data and real data from TerraSAR-X satellite corresponding to a building—North Star Times Tower in Beijing, China. Finally, we compared the signal detecting based methods involved in this paper, with the compressed sensing (CS) methods (e.g., regularized orthogonal matching pursuit (ROMP) and L1-norm regularization). The results demonstrate that the CS methods achieve only the spatial information of the scatterers, whereas the signal detecting based methods are capable of obtaining the backward scattering power information of each scatterer in parallel with the spatial information. This leads to the conclusion that the signal detecting based methods gain more information about the scatterer than the CS-based methods. Furthermore, the detection accuracy and time cost of the aforementioned methods are presented and discussed.

In this study, the optical system of 64×64 InGaAs/InP avalanche photodiode (APD) focal plane arrays for single photon detectionin in the near-infrared region (0.9~1.65 μm) was designed. The optical system of APD arrays is of great importance, since APD is much more sensitive to light compared with PD detectors, even a slight optical system change may affect the detector working state greatly. The optical design of the window, MLA and SPAD were focused in this study. The window is a typical glass-to-metal structure consist of a metal base part (4J29 alloy) and a glass window part (saphhire). The surface of the 4J29 kovar alloy is coated with Ni with the sickness from 1.30 μm to 8.90 μm, and the surface of the Ni coated layer is over lapped with another Au layer (sickness: 1.30 μm to 5.70 μm). In order to improve the light absorption efficiency, micro lens array (MLA) was used to focus the light to each pixel. Proper optical structure in the SPAD could suppress optical crosstalk, by adding optical isolate structure and a anti-reflection (AR) coating layers in SPAD.

In order to study the lattice damage caused by low-energy argon ions on the single crystal silicon substrate and the influence of incident energy on the substrate damage during the ion beam polishing process, the molecular dynamics (MD) was used to simulate the incidence of a single argon ion on the single crystal silicon substrate. Compare the lattice defects produced by the bombardment of single crystal silicon by ions of different incident energy. Ion beams with different incident energy were used to polish the surface of single crystal silicon, and the surface roughness after polishing was compared. Experimental and simulation results show that ion bombardment will cause lattice damage to the surface of the substrate. As the energy increases, the range of lattice damage will expand and the resulting lattice defects will be more dispersed. The use of low-energy (200~600eV) ion beams can further reduce the surface roughness of the substrate on the basis of reducing lattice damage

In this paper, the effects of uncertain factors in the thermistor pasting technology are researched using parameter sensitivity analysis method. The relevance between thickness of electrical insulating tape, contact interface area, thermal contact conductance and thermal time constant is analyzed, and sensitivity classification of uncertain factors is done by the absolute/relative variation of thermal time constant. Research shows that 1) thickness of electrical insulating tape d_PI and thermal contact conductance h_t are the insensitive parameters about thermal time constant in the surface pasting technology, relative sensitivity indexes are 0.169[s^0.5/μm^0.5] and 0.292[m·s^0.5·K^0.5/W^0.5] respectively; 2) contact interface area At is the sensitive parameter about thermal time constant in the surface pasting technology, absolute variation and relative sensitivity index is 124.6s and 3.530[s^0.5/mm] separately; 3) in addition to the high-precision calibration, filtering the sensitive-bead size and contact interface area strictly are also critical for improving the real-time measurement accuracy and temperature-changing response speed of thermistors.

In this work, the laser-induced periodical surface structure (LIPSS) on silicon is generated by 532nm nanosecond laser with pulse duration of 10ns and repetition frequency of 10Hz. The formation of LIPSS is observed under laser fluence of 260mJ/cm² and pulse number of 600 shots with p- and s-polarized laser. For p- polarized laser, when the incident angle is 10°, 20°, 30°, 45°, and 60°, the LIPSS period is 400nm, 743nm, 902nm, 1148nm and 2525nm correspondingly. Besides, the orientation of LIPSS is always perpendicular to the polarization direction of the incident laser. For s- polarized laser, the LIPSS is perpendicular to the polarization direction and the period increases with the incident angle when it is less than 42°. The period is 457nm, 515nm, 549nm and 610nm at incident angle of 10°, 20°, 35°, and 41° correspondingly. when the incident angle exceeds 42°, a set of crossed LIPSSs emerge which are symmetric about the polarization direction. Meanwhile, the period reduces a lot which is around 400nm.

In the field of Ground-based Interferometric Synthetic Aperture Radar (GB-InSAR), permanent scatterer (PS) is mainly used as the high coherent point, and its selection method widely uses the amplitude dispersion method and the correlation coefficient threshold method. In general, the highly coherent points selected by these two methods can meet the deformation measurement requirements. However, in some areas with complex environments (such as vegetation growth, rainfall and snowfall), the quantity and quality of high-coherent points selected based on these two methods cannot meet the requirements of deformation monitoring. Aiming at the problem that the number of selected highly coherent points is small or the quality is not high in a complex environment, the article proposes an improved highly coherent points detection method: the method uses non-local means filtering to reduce image noise, and then combines the amplitude and phase information of the image with the fuzzy c-means (FCM) clustering algorithm for high-coherent points selection. Finally, through the simulation of measured data and the simulation results show: for the complex environment measured in the article, compared with traditional methods, more high-coherent points that meet the requirements can be selected by using improved methods.

For the problem of the detection and tracking about multiple maneuvering fluctuating weak targets in radar systems, a track-before-detect algorithm based on interactive multi-model multi-Bernoulli filtering (IMM-MBTBD) is proposed. In a track-before-detect algorithm, the squared modulus likelihood ratio (SLR) is usually utilized to match the possible trajectories of targets. In this method, however, only the amplitude information is considered and the phase information is ignored, which leads to a decrease in filter performance. In this paper, we use the complex likelihood ratio (CLR) instead of the SLR in the IMM-MB-TBD and preserve the spatial coherence of the phase to realize the detection and tracking about multiple amplitude fluctuation targets with the Swerling 0,1, and 3 fluctuations. Furthermore, in order to accommodate to the situation where the prior information of target births is unknown and overcome the difficulty of detecting a weaker target and a stronger target at the same time when the targets fluctuate, an adaptive birth algorithm based on measurement likelihood ratio is proposed. Simulation results show that at low signal-to-noise ratio, the proposed IMM-MBTBD algorithm provides better performance in estimating of the state and number of targets.

InAs_0.92Sb_0.08 and InAs_0.06Sb_0.94 epilayers have been grown on InAs substrates by liquid phase epitaxy (LPE) and melt epitaxy (ME) respectively. InAsSb epilayers grown by LPE and ME have been firstly observed by a high temperature laser scanning confocal microscopy (LSCM) . X-ray diffraction (XRD) measurements confirmed that the lattice constant of InAs_0.92Sb_0.08 epilayers grown by LPE is 6.0912 Å, while that of InAs_0.06Sb_0.94 epilayers grown by ME is 6.4520 Å. Uncooled photoconductors were fabricated using InAs0.06Sb0.94 thick epilayers grown by ME. The photoresponse wavelength range of InAs_0.06Sb_0.94 detectorsis 2-10 μm. At 295 K, the peak detectivity D_λp* is 3.59 × 10⁹ cm Hz^1/2W^-1 at the wavelength of 6.5 μm, and the detectivity D* is 1.0 × 10⁹ , 2.1 × 10⁸ and 1.0 × 10⁸ cm Hz^1/2W^-1 at 8, 9 and 10 μm for the detectors with Ge immersion lenses respectively

In the flight test of the airborne airdrop system for transport aircraft, it is necessary to measure the operating speed and trajectory of the airdropped cargo in the cabin to verify whether the function and performance of the airborne airdrop system for aircraft meets the requirements of combat wood targets. According to the airdrop environment and mission characteristics, this paper constructs a spatial positioning model based on the time difference of arrival. The crosscorrelation interference and non-line-of-sight effect elimination technology is adopted to achieve rapid capture of weak signals and improve the accuracy of tag positioning. The position accuracy of the test system based on UWB is better than 0.15 meters, and the frame rate is 25 frames. This method can also be applied to the measurement of suspension and other tasks such as hanging swing measurement.

The space flexible solar wing is easy to be affected by the vibration from the satellite platform, which has characteristics of low frequency, dense frequencies and large static deformation. In order to ensure that the extension mechanism of the large flexible solar wing can realize the reliable, repetitive and stable deployment and retraction of the sun tracking, adynamic model of the flexible solar wing is established. On this basis, a double closed-loop control strategy of improved sliding mode and variable integral PID regulator is proposed to realize the double closed-loop control system of speed and current. Finally, the experimental system of peripheral circuit, driving circuit and position detection circuit based on single chip anti fuse FPGA is designed, and the experimental verification of large flexible solar wing deployment control based on FPGA is realized. Through the experimental verification, the system can achieve the vibration suppression function of the extension mechanism with high disturbance, and realize the engineering application that the large flexible solar wing can be repeatedly and smoothly unfolded and folded which has a certain application value and provides a reference for the subsequent engineering application.

In aerospace and high power laser devices, fused quartz material is in large demand because of its strong adaptability.The fused quartz optical mirror blank was processed by ultra-precision grinding, magnetorheological machining (MRF) and small smooth polishing (SP).Ultra-precision grinding machine was used for rough pre-polishing to obtain high precision surface shape, and then MRF was used for ultra-precision polishing to further improve the surface shape.The medium and high frequency errors generated by MRF processing were controlled by using smooth polishing.The experiments of ultra-precision machining of fused quartz mirror were carried out by using the combined process, and the rapid convergence of deterministic machining and full frequency error was realized.The combined process can not only improve the processing efficiency of optical mirror, but also reduce the processing cost and realize short process processing.

At present, most of the deep learning target detection methods based on multimodal information fusion are integrated, which makes the fusion image quality cannot be directly controlled. It is not conducive to strengthening the target detection of the network in principle. A multimodal information fusion detection method based on generative countermeasure network (FusionGAN-Detection) is proposed, which is composed of GAN and a target detection network. Aiming at the uncontrollability and blindness of existing information fusion detection algorithms, the new method introduces generative countermeasure network for information fusion. It uses loss function and dual discriminator to provide controllable guidance for generator. In the process of information fusion using GAN, the loss function of traditional thought can extract the information which is beneficial to target recognition to the maximum extent, and avoid the loss of channel information. The detector acts as a discriminator during the training process to guide image fusion and promote the improvement of image quality. Meanwhile, it acts as a target detector for target detection during testing. In order to verify the effectiveness of the method, KITTI data sets are used for training and testing. The experimental results show that new method is better than the existing advanced methods in AP.

We propose a feature fusion method for gesture recognition based on Frequency Modulated Continuous Wave (FMCW) radar. First, we estimate the radial information distance, Doppler and tangential information azimuth, elevation of the gesture by signal processing to construct the multi-dimensional feature data set; Then, for feature extraction and accurate classification, we design feature fusion scheme and build multi-dimensional feature convolutional neural network. Experimental results show that our proposed scheme with gesture multi-dimensional feature as input can solve the problem of insufficient feature representation in traditional Range-Doppler (RD) domain gesture recognition methods, and the recognition accuracy is improved by 4%~8% compared with the case without feature fusion.

Differential Tomography SAR (D-TomoSAR) is a significant technology for large-scale land subsidence monitoring, which can be utilized to obtain deformation information. In particular, it owns unique technical advantages, such as high deformation sensitivity and accuracy. D-TomoSAR allows resolving layovers by handle the situation where multiple scatterers exist within a range-azimuth resolution cell. However, the extraction of scatterers is a critical and challenging issue. As we all know, there are four main parameters for a PS in SAR images. They are namely amplitude, phase, coherence and signal-to-clutter ratio (SCR). In this paper, the characteristic of the four parameters is analyzed firstly. Then, according to the characteristic of the four parameters, three different methods of PS extraction are studied theoretically, which are amplitude dispersion (AD) threshold method, coherence coefficient (CC) threshold method and SCR threshold method in temporal and spatial domain. Finally, the SAR data from TerraSAR-X satellite are used for experimental at North Star Times Tower in Beijing, China. The extraction results of the three approaches are compared. The experimental results show that AD and SCR methods are more reliable than CC method. It is due to that CC approach is affected by the neighborhood pixels to a large extent. Additionally, SCR method is more time saving than AD and CC methods. Therefore, we can conclude that SCR approach is suitable for engineering application.

As an important application of computer vision, visual tracking has being a fundamental topic. Compared with visible image, infrared image has the characteristic of low resolution, blurred contour and single color feature. Thus, it is still a challenge for infrared object tracking. Further, it is difficult to balance the real-time performance and accuracy. This paper proposed a tracker based on the SiamRPN tracker with a deeper and lightweight MobileNet V2 structure as the backbone network. During network training, the weights are updated by the scheme of the model-independent metalearning method. The computed model only passes a few gradient descents on the first frame can obtains the most suitable for the current frame. In the end, the tracker is tested on various datasets. Experimental results show that the tracker can achieve a balance between tracking accuracy and inference speed, which is crucial for deployment on mobile devices.

The total impulse of the disc type laser ablation thruster is linearly related to the number of ablation points on the target plate and improving the utilization rate of the target plate is beneficial to obtain more ablation points on the limited target plate surface. In this paper, the structure of the laser ablation micro-thruster was designed, two ablation paths for thruster were put forward and analyzed, namely the circle path and the spiral path. The distribution of the ablation point on the target plate was theoretically modeled. Through calculation, the influence law on the utilization rate of the target plate by the size of the target plate and the size of the ablation point under the two paths is analyzed and compared. The results show that, the target utilization theory of circle path and spiral path can reach up to 78.54%; The research in this paper provides theoretical guidance and design reference for the utilization of the target plate of the disc laser ablation thruster.

Aiming at the problem of abnormal sound of lens caused by incomplete tightening of pressure coil after optical lens installation, this paper presents a method for calculating and verifying the tightening moment of pressure coil. According to the mechanical dimension of the screw thread of the pressing ring, the mechanical properties of the pressing ring of the mirror tube and the optimal number of turning rings were calculated and analyzed. In order to meet the practical application requirements, the tightening moment of the pressure coil obtained by the solution was brought into the 3D model for simulation, and the finite element modeling analysis of the tightening moment of the pressure coil of the optical lens was completed, so as to further solve the displacement variation of the lens group. Finally, the above analysis results were quantified by torque wrenches, and random vibration was carried out at one third of the order of acceptance level vibration test conditions, so as to complete the bottom analysis and verify the tightening torque of the screw ring under different ratios of XM-31 adhesive and vulcanizing agent. The experimental results show that the characteristic sweep curve of the optical lens does not change before and after the vibration, the mode of the optical lens remains stable all the time, and the data modal analysis results are consistent with the actual situation.

In the conventional infrared-polarization fusion method based on HSV color space, the problem of low color contrast and information loss or artifacts always exist which would impact the observation and processing of the fusion results. The cause of low color contrast in conventional method is analyzed in this paper, and an improved method is proposed. We find that some improvements could be done based on Channel S and Cannel V separately which could enhance the details in the regions of interest and maintain the precision of image as well. For Channel S, saliency extraction and morphological filtering are implemented to obtain the salient features of intensity image and degree of polarization image, while the redundant background information could be suppressed. The detail enhancement in Channel S could be realized through the fusion of common and unique feature of the salient images extracted above. For Channel V, the common and unique feature fusion between the salient degree of polarization image and intensity image could enhance the details in Channel V. Experimental results have proved that the method we illustrate could obtain better color contrast and less artifacts compared to the conventional fusion methods.

The material removal mechanism of aluminate magnesium spinel was investigated by nanoindentation and microscratch test. The hardness increases with the increase of load. The influence of scratch speed and scratch load on the removal mechanism of spinel was analyzed based on the penetration depth, tangential force and acoustic emission (AE) signal. It was found that with the increase of scratch speed, the area of elastic-plastic removal and brittle plastic transition increased. The AE signal appears at the beginning of a brittle plastic transition for the first time. Moreover, the tangential force growth rate of spinel in the elastic-plastic stage is lower than that in the brittle removal stage. The results indicated that improving scratch speed and reducing the scratch load is beneficial to reduce the range of brittle removal of spinel, thus increasing the range of plastic removal and brittle plastic transition.

In order to realize the weak signal accurate imaging of ultra-long range target, a single photon detection imaging system is established. The detection technology of neighborhood second-order spatiotemporal correlation is proposed, and the detection probability and false alarm probability of neighborhood second-order spatiotemporal correlation algorithm are analyzed. The simulation analysis of the neighborhood second-order spatiotemporal correlation algorithm shows that for more than 90% detection success rate, the average photon number required by the conventional method is 0.77, which is reduced to 0.0375. The results show that it can basically meet the requirements of high imaging accuracy, low false alarm rate and fast imaging speed for ultra-long range targets.

A compact micro-integrated narrow-linewidth extended cavity diode laser (ECDL) based on a volume holographic Bragg grating (VHBG) around 780nm is demonstrated. The ECDL consists of gain chip, micro-optical lenses, VHBG, optical isolator and thermistor. All components are integrated on a AlN ceramic plate, which is on the top of a thermoelectric cooler (TEC). Moveable parts are strictly omitted to provide excellent mechanical stability. The ECDL provides a short-time linewidth(20us) of 70 kHz with 30mW optical power. The mode-hop free range of the ECDL is up to 9GHz. The laser frequency variation is ±1MHz in 40min. The ECDL also has the advantage of the compactness of a standard 14-pin butterfly package with a footprint only 30 mm×12.7 mm.

Mid-wave infrared imaging has the advantages of all day, high resolution, strong adaptability to the environment, so it is widely used. In high-speed target tracking and aiming. The target flying speed is extremely fast and the DOF (depth of field) of the fixed focal lens is limited, so the zoom lens is required. In order to ensure that the tracking target will not be lost, a fast zoom requirement of no more than 0.3s is proposed. According to the requirements, a compact athermalized mid-wave infrared dual field fast zoom lens is designed. The focal length is 400 mm and 200 mm, and the passive athermalization design is adopted which can meet the working temperature of - 40° to + 60°. the lens can match the cooling detector with effective pixel of 640 × 512, size of 15 μm and F/4 .The volume of the detector can be controlled in the range of L (196mm) × w (116mm) × H (185mm). The zoom group structure is considered in the design, and the switching time is less than 0.12s, and the image plane is stable when the zoom group switching.

Using first-principles calculations, the terahertz dielectric and thermal properties of cubic silicon carbide (3C-SiC) are investigated for a wide range of temperature ranging from 0K to 2500K with the density functional perturbation theory (DFPT) and the quasi-harmonic approximation (QHA).

The optical on-chip integration technology can be used to realize spectrum analysis. Compared with the traditional bulky high performance optical spectrum analyzer which is always used in laboratory, on-chip spectrometer has the advantages of chip-scaled, low-cost and suitable for detection in complicated environment. For example, it can be utilized for detection of toxic gases such as carbon monoxide and hydrogen fluoride. In the past decade, several kinds of on-chip spectrometer, dispersive spectrometer and Fourier transform spectrometers, have been demonstrated as promising candidates for wide range of spectral application. In this paper, a digital Fourier transform spectrometer based on an interferometer whose arms consist of several optical switches which are connected by Mach-Zechnder Interferomter (MZI) with different arm lengths is demonstrated. The optical switch based on thermo-optic effect is used to select the different arms of an MZI, so the combination of optical switches and MZIs can lead to a series of different optical path difference. The proposed on-chip spectrometer is designed for Original-band (O-band) and the prospective spectral resolution is 0.3 nm. The future effort will focus on the test of the proposed spectrometer and the followed data processing including dispersive compensation and spectral reconstruction.

One of the main sources of fiber optic gyro noise is the polarization error of gyro. Integrated fiber optic gyroscope can reduce gyroscope volume and cost. Many researchers at home and abroad have designed and produced integrated optical path devices of fiber optic gyroscope, but there is still a lack of an effective performance evaluation method. According to the physical model of fiber optic gyroscope and Jones matrix theory, the optical devices in small fiber optic gyroscope are studied in this paper. The physical model of discrete optical path device and the transmission model of integrated optical path device system of small interferometric fiber optic gyroscope are established. The influence of non-ideal factors of integrated devices on polarization error of gyro system is analyzed. In the design process of integrated optical path gyroscope, the research method in this paper can select the components according to the different precision requirements of gyroscope. The research method in this paper can effectively evaluate the performance of integrated optical path devices of small interferometric fiber optic gyroscopes.

Unknown wall parameters will cause the target position offset and image distortion for three-dimensional imaging of through-the-wall radar. In this paper, we propose an autofocusing imaging method of alternating learning of wall parameters and sparse coefficients based on the parametric diffraction tomography sparse model. The sparse coefficients reconstruction algorithm is unrolled a multilayer neural network for learning to update the sparse coefficients. The normalized mean square error is selected as the network loss function, and the batch gradient descent method is used to learn the network parameters. Subsequently, we still use this neural network to learn the variation of wall parameters to update the wall parameters, but the difference is to use the image quality evaluation function as the loss function. After several alternating iterations, we determine the sparse coefficients and wall parameters that minimize the loss function as the final result. The simulation results show that this proposed method can effectively eliminate the target position offset and image distortion, so that it can realize accurate estimation of wall parameters and autofocusing imaging.

This paper presents focus on physical feature extraction and parameter estimation for satellites in space. In the context of increasingly prominent space security issues, it is necessary to improve the security and protection capabilities of the space system, especially the refined feature extraction and recognition technology of space targets. However, the feature extraction methods based on optical or radar imagery are limited by resolution and observation conditions, making it difficult to extract refined features. This paper is based on the attribute scattering center theory, using high-frequency electromagnetic wave detection method. Through the attribute parameter estimation result of the scattering center, the scattering structure on the space target can be identified and its distribution position can be analyzed, and the structure and size of the typical parts in the space target can be extracted. Aiming at the problem of high parameter dimension and large amount of calculation, an attributed scattering centers parameter estimation method with amplitude and phase separation is proposed, which reduces the time complexity of parameter estimation by an order of magnitude, saves system resources and increases the parameter estimation rate, and the estimation accuracy is better than 90%. In terms of the efficiency and accuracy of parameter estimation from simulation and experimental data, the effectiveness of the proposed method in this paper is verified.

Most of the traditional compressed sensing ground penetrating radar imaging is performed in a two-dimensional form, and it is impossible to accurately determine the specific location of the underground target. In order to solve these problems, a three-dimensional imaging method is proposed for ground penetrating radar. In this method, the data is observed from a planar MIMO array, and a dictionary matrix is constructed into a three-dimensional model. Finally, the underground scene in three dimensions is reconstructed with the total variation minimization. The experimental results show that in the layered media scene, the proposed method can accurately image the target by using only a small amount of data.

With the wide application of UAV (Unmanned Aerial Vehicle) in social life, the threat to social public security is increasing, and UAV countermeasures technology has gradually become a research hotspot. Firstly, according to the working process, UAV countermeasures can be divided into three types: detection and early warning technology, signal interference technology and interception and strike technology, and their principles are expounded. Finally, the advantages and disadvantages of countermeasures technology are analyzed. The problems existing in UAV countermeasures are analyzed from three perspectives of integration, collaboration and innovation, and the future construction of countermeasures is prospected.

A novel compact microstrip resonant cell (CMRC) with an ultra-wide suppressed frequency band is proposed. It consists of a pair of symmetrical stepped trapezoidal-shaped patches, two cross-connected high impedance microstrip lines and two low impedance open-ended microstrip stubs. A demonstrative low-pass filter (LPF) using a single cell of the stepped trapezoidal CMRC (STCMRC), with a compact size of 23.8mm×11mm, is designed, fabricated and measured. The insertion loss (IL) including that of testing fixture is less than 0.37dB in the pass band of 0 to 2GHz. The suppressed frequency band with a rejection better than 10dB spans from 3 to 18.2GHz, corresponding to a relative frequency bandwidth of 143%.

Surface reconstruction method plays an important role in many engineering fields. It is an imperative procedure to carry out surface reconstruction from measurement data in reverse engineering, which is complicated with the presence of outliers. To achieve better accuracy and robustness of reconstruction, an improved moving total least squares (MTLS) algorithm based on k-means clustering called KMTLS method is proposed in this article. KMTLS adjusts the weights of discrete points within the support domain by adopting a two-step fitting procedure. Firstly, ordinary least squares (OLS) method is adopted to obtain the pre-fitting result and calculate the residuals as the input of k-means clustering. In kmeans clustering, abnormal nodes are classified into one cluster and a weight function based on clustering information is introduced to deal with these nodes. Secondly, based on the compact weight function in MTLS and the weight obtained in the pre-fitting procedure, weighted total least squares method is conducted to determine the final estimated value. The process of detecting outliers is automatic without setting threshold artificially. The experiment shows that KMTLS has great robustness to outliers.

Simulation of rocket plume infrared images by computer programs is an important method to study their infrared radiation characteristics. In order to improve the computation efficiency, a simulation method based on CUDA was proposed. Firstly, the characteristics of the LOS (Line of Sight) method in radiative transfer calculation were analyzed. All the path points of LOS method were calculated in advance, and the parallel level was promoted to the number of path points of LOS method and bands of spectrum. In this case, more threads could be mobilized at the same time, and the utilization of GPU was improved. Secondly, A hardware bilinear interpolation method using CUDA texture memory was proposed, which greatly improved the calculation efficiency of physical parameters of the components. Finally, CUDA was also used for acceleration in the projection imaging module of spaceborne infrared sensor. The simulation results show that using CUDA for parallel computing to realize physical parameters search and projection imaging can greatly improve the overall simulation efficiency of rocket plume infrared images.

In order to solve the problems of limited computing resources on the fuselage and autonomous recognition of visual images during the real-time control of the flapping-wing flying robot, and to realize the intelligent navigation control of the flapping-wing flying robot based on vision, this paper proposes a visual image processing method and edge detection method for a flapping-wing flying robot based on cluster analysis. Using cluster analysis method to classify the visual image pixel data of the flapping-wing flying robot, and using the classic edge detection algorithm to perform edge detection on the flying robot image after clustering analysis, and reasonably extracting the visual target contour features and environmental edges of the flapping-wing flying robot feature. An experimental platform for a flapping-wing flying robot was built with a flapping-wing flying robot and an airborne vision camera. Flight experiments are carried out on the visual image processing method and image edge detection method of flying robot based on cluster analysis. The experimental results verify the feasibility and effectiveness of the method in this paper. The innovation of this paper is embodied in the application of cluster analysis method to the visual image processing of the flapping-wing flying robot, and the edge detection of the visual image of the flapping-wing flying robot.

With the development of laser technology, laser drilling has been widely used in industrial processing, especially the manufacture of micro-hole arrays on metal or non-metal surfaces to improve the surface properties of materials. The manufacture of laser micro-hole arrays depends on high-frequency laser and high-speed scanning devices. Compared with the traditional galvanometer scanning device, the scanning speed of the polygon mirror scanner can reach 1000 m/s, which is dozens or even hundreds of times that of the galvanometer. Polygon mirror scanner combined with the high repetition rate nanosecond laser can realize the ultra-high-speed two-dimensional plane precision micro-hole arrays. High-speed laser precision drilling, especially micro-hole manufacturing, has become the core technology of high-speed laser micro-machining. Based on the FPGA control system, the scanning speed of the polygon mirror scanner can be synchronized with the pulse repetition rate of the nanosecond laser, and the dot-arrays drilling can be effectively realized. There is only one laser pulse injection at each scanning position. This paper mainly studies the effects of average laser power, laser pulse repetition rate, laser pulse width, and working defocus on laser drilling. Including hole diameter, the conical degree of each hole, and drilling efficiency. The hole diameter and the threshold energy density of material burning under different laser pulse widths were tested and compared. The changes of laser drilling aperture and taper were observed with increased laser power and single pulse energy. While the laser pulse width is 120 ns, the material burning threshold is 4.59 J/cm², and the maximum hole radius is 50 μm. The minimum taper is 0.36. The optimized processing technology is proposed. Under the same other conditions, the hole depth and hole morphology of 0.3 mm stainless steel dot-arrays arrangement drilling, dot-arrays dislocation arrangement drilling, and dot-arrays dislocation arrangement drilling under zoom were experimentally studied. While the number of scan times is 40, the maximum hole depth of the dot-arrays staggered arrangement zoom drilling can reach 161 μm.

The space data transmission antenna system is an important channel for the telemetry data transmission of the satellite platform to the ground. Therefore, the logarithmic antenna has high requirements for the pointing stability to the ground. The pointing accuracy of the data transmission antenna is seriously affected by the vibration of the satellite platform. Therefore, this paper analyzes the vibration of an orbiting spaceborne data transmission antenna, models, simulates, analyzes and calculates the logarithmic transmission antenna, optimizes the antenna system structure based on this, and designs a fuzzy adaptive PID control system under decentralized control for closed-loop control, The simulation test of the antenna system under the real vibration environment is carried out through the vibration test. The results show that the antenna system has stable performance and can effectively suppress the vibration, which verifies the reliability and feasibility of the antenna system. At present, the data transmission antenna has been put into orbit with satellite launch, and its performance indexes are stable, the research results can be used as a reference to improve the vibration safety of relevant spaceborne antennas.

In view of the urgent demand high resolution and large field of view on three-line array airborne mapping camera, the paper proposes a new idea which applies monocentric multiscale principle to three-line array airborne mapping camera with high resolution and large field of view (FOV), and gives the design method and design process of optical system. monocentric multiscale optical system is designed, focal length is 70mm, resolution is 0.1m@2Km, and there are four multispectral bands which are panchromatic、R、G、B band. The Base to Height ratio (B/H) is 0.83. The FOV of monocentric multiscale optical system is 80°, and working FOV is 60° which can be extended to 105°. The optical system has excellent imaging quality and successfully solved contradiction between large FOV and high resolution.

Aiming at the problem of insufficient similarity between the binary defocus fringe generated by the traditional dithering algorithm in the binary surface structured light coding technology and the standard sinusoidal fringe, a dynamic threshold dithering algorithm based on the traditional Floyd-Steinberg dithering algorithm is proposed, which can generate the binary fringe pattern with high sinusoidal similarity, improving the accuracy of structured light three-dimensional reconstruction. Considering the influence of the pixel gray value in the vicinity of the pixel point on the threshold of the dithering algorithm, the parameters α and edge sensitivity coefficient 𝓂 are introduced to dynamically adjust the dithering threshold to improve the sinusoidal similarity of the dithering stripes. In addition, a total error calculation and evaluation method that takes into account the effects of gray error and phase error is used to improve the robustness of the algorithm. The simulation shows that when the parameters α and 𝓂 in the algorithm are set to 0.68 and 0.20 respectively, the total error is reduced by 82.97% at the fringe period T = 20 pixel. Corresponding experiments have been established to verify that the dynamic threshold can be displayed more clearly in the small change area of the model, which improves the image quality of the three-dimensional reconstructed object.

In order to study the spatial distribution and temporal variation of land desertification in the four counties of Bashang, Zhangjiakou, Hebei, a total of two time-phases of Landsat-8 remote sensing data in 2017 and 2018 were used as data sources, taking Normalized Vegetation Index (NDVI) as the main quantitative remote sensing interpretation method to derive the patterns of land desertification changes in the four counties of Bashang, Zhangjiakou, Hebei in recent years .The results showed that from 2017 to 2021, the chang of lightly and moderately desertified land areas in the four counties of Bashang, Hebei were generally stable, the areas of heavy and extremely desertified land decreased significantly, the condition of overall land desertification has improved significantly.

This paper reports a liquid optical analogical gravitational lensing made on a chip. It concludes a mixing liquid stream as a medium and a PDMS chip as a mixing cell. Comparing to other optical analogical gravitational lensing, the liquid lens uses mixed liquid to realize the non-uniform change of the medium. It is simple to manufacture, low cost, small size and easy to preserve. Changing the velocity and ratio of liquid stream will adjust the refractive index (RI) profile of the entire waveguide. It helps to visually show the effects of gravitational field on light.

In this study, the package structure of 64×64 InGaAs/InP single photon avalanche diode (SPAD) was analyzed and designed. The vacuum package shell which contains 104 pins is made of kovar alloy (4J29). The shell window is a typical glass-to-metal structure, which consist of a metal base part and a glass window part (sapphire). W-Cu heat sink was used for the extraction of heat generated by the chips and the thermoelectric cooler (TEC). Micro lens array (MLA) was used to concentrate the light to improve the light absorption efficiency. The chips were mounted on a AlN ceramic substrate, which is used as a wire bonding buffer layer, as well as a good heat dissipation channel. A three-stage TEC was used to stabilize the SPAD working temperature at about 210K. Schematic diagram of the wire bonding designs, and the main technological process steps of package were also given.

Object detection technology has always been one of the important research directions in the field of computer vision. It has important application value and prospects in both civil fields and military fields. With the emergence of artificial intelligence technology, deep learning has gradually replaced the traditional algorithm with its higher accuracy. Considering that most of the current algorithms are for color image, compared with color image, infrared image contains less feature information, and it is more difficult in object detection. In this paper, different algorithms are used for object detection in infrared images, and the detection results are compared. This paper chooses YOLO V5 and combines it with MobileNet to lighten the model. After lightening, the parameters are reduced by 30%, but the accuracy is only reduced by 5%. Finally, this paper quantify YOLO V5 based on the model quantization method of PyTorch. After quantization, the accuracy of the model decreases by 2%.

An elaborate Hadamard transform infrared imaging spectrometer system based on the digital mirror device (DMD) is proposed in this paper, through which the implementation of Hadamard coding in this system is revealed. Derivation of the process of spectra radiation transmission covers the influence factors of target radiative characteristics, atmospheric environment, optical system, dispersion, DMD modulation, spectrum response of detector, Hadamard encoding and decoding etc. A simple physical scene is introduced including diverse materials and model tank and simulation results are given by the transfer model with real-time meteorological data, real-time temperature data and imaging system parameters.

In the antenna array structure of passive radar seeker, uniform circular array is usually selected. However, due to the influence of spatial distribution, missile structure and radome design constraints, the non-uniform circular array antenna array structure is needed in special cases. In this paper, the principle of direction finding ambiguity generation of multiple signal classification is described, the solution design of the maximum radius of the uniform circular array and the non-uniform circular array is analyzed, at the same time, for the non-uniform circular array structure, the ambiguity performance of the array under the maximum radius and greater than the maximum radius is compared and analyzed, the effectiveness of the theory in the nonuniform circular array is verified by simulation. This paper has a reference value for solving the ambiguity of direction finding under the arrangement of passive non-uniform circular array in engineering application.

In order to meet the needs of TV detection in deep sea environment, a compact continuous zoom system driven by cam sleeve is studied. Through the analysis of the system indicators, the optical system of three groups of linkage zoom and rear group for focusing is selected, and the structure of cylindrical cam driving to cooperate with the cylindrical guide sliding is clarified. Based on this idea, each module of the TV detection system is designed in detail. The design method of electric iris diaphragm in the middle of the system is proposed, and the calculation process and results of zoom drive mechanism selection are given. The method of alignment of lens optical axis adjustment with theodolite and cross-reticle is put forward. With this design and adjustment process, the TV detection system can be applied to deep sea environment is finally realized. The system has a continuously variable field of view with 5.8°~60° underwater, the F number changes from 3.5 to 16, and the whole zoom time is less than 2s. The test results show that the MTF of the center field of view at the long focal point and the short focal point are 0.4 and 0.57. Test images in laboratory and underwater scene are clear and high resolution, which shows that the imaging quality of the system is excellent. At the same time, the system is Φ 105×115mm in actual size and 860g in weight, and can be integrated into the deep-sea pressure cabin to meet the needs of deep-sea detection.

Metamaterials have unlocked a number of unique optical properties for manipulating the electric and magnetic dipole resonances have drawn much attention over the past decades. It has been recognized that a metamaterial could provide a new approach to excite electromagnetic behaviors leading to potential applications in the fields of sensing, energy, and information The low-loss conformal processing integration of multi-band wide-angle electromagnetic antireflection metamaterials, functional layers and radome is a challenging research. In this paper, a high-permeability metamaterial with similar electromagnetically induced transparency (EIT) characteristics is constructed around the metal split ring, and the equivalent dielectric constant and equivalent permeability of the split ring meta-unit are analyzed based on the equivalent medium parameter inversion model. Analyze the regulation mechanism of array structure parameters on electromagnetic characteristics. The research results show that this EIT wave-transmitting material has the characteristics of wide working frequency, high transmission rate, and insensitive to electromagnetic wave incident angle and polarization direction. In view of this, a multi-band broadband electromagnetic antireflection metamaterial with a multi-layer nested coupling structure (open resonant ring-chiral spiral ring-cross symmetrical structural unit) is designed, which is continuous in different polarized electromagnetic waves from 0 degrees to 80 degrees. Large-angle oblique incidence 11GHz, 15GHz, 33Ghz three-band high transmission rate, bandwidth up to 1.3GHz, 1.02GHz and 1.678GHz.

In this paper, we design a feature fusion module for multi-model imaging based on deep learning. The fused feature of multi-dimensional images is used for object detection, which can effectively avoid interference caused by complex environments. Feature fusion module consists of a convolution layer and an activation function. It can establish the connection between different images. The fusion rules are obtained through supervised learning. Compared with the traditional target detection structure, it can extract more detailed information from several source images. Feature maps extracted from each image are fused by the feature fusion module to form a new feature map. Such a feature map can be better used for the generation of objects masks and bounding boxes. We capture a series of multi-dimensional images with a flexible multi-model camera. When shooting, multi-dimensional information is simultaneously recorded in an image. Through decoding, multiple images of different types can be obtained, including polarization and spectral images. These images record the multi-dimensional optical characteristics of the object and background. Compared with the traditional single-input color or monochrome image method, the proposed method gets 0.25 of average precision and 0.75 of F1-score, which achieves higher detection accuracy in various natural backgrounds.

Multiple elevation beams (MEBs) mode is a new kind of operation mode for high-resolution wide-swath spaceborne synthetic aperture radar (SAR) imaging. This mode uses a single wide transmit beam to cover the whole swath in range dimension; while on receive it forms multiple narrow scanning beams with high gain to receive echoes from multiple subswaths simultaneously. Thus, one of the key techniques associated with this mode is to extract data of each subswath from the received overlapping echoes. To deal with this issue, this paper brings forward an improved two-step data separation approach which combines the peak forming of antenna beam pattern onboard and nulls forming on ground. This approach has the advantages in the following three aspects. First, it can guarantee that echoes of each subswath will be received with optimal antenna gain without any loss; second, it can accommodate the effect of beam mispointing due to the error in the digital elevation model (DEM) and thus has good robustness in nulls-forming; last, it transfers part of the beam-forming process from space segment to ground segment so as to reduce the processing complexity on satellite, thereby improving its feasibility as well as precision.

In infrared imaging system, nonuniformity is the key factor limiting the improvement of imaging quality. In this paper, an adaptive neural network nonuniformity correction algorithm based on motion detection is proposed. The motion estimation algorithm based on gray projection is used to select the learning reference frame of neural network algorithm. Combined with edge detection algorithm and neighborhood variance information, the learning step size of neural network is adjusted adaptively, it effectively solves the “ghost effect” caused by insufficient scene motion in the traditional neural network algorithm. The algorithm automatically extracts effective image frames from the actual scene and updates the correction parameters, without planning the correction area and scene motion mode, and realizes automatic real-time correction.

We have used Terahertz time-domain spectroscopy(TDS) to investigate the dielectric properties of water and water solutions of containing different ions. The complex permittivity of NaCl solutions, for different concentrations of ions, was obtained by attenuated total reflection(ATR) method. We have observed that, when increasing o ion concentration, the real part and imaginary part of the permittivity are increased. Furthermore, we also calculate the dielectric constant of ionic aqueous solutions with different concentrations based on molecular dynamics method. The simulation method, based on Newton mechanics, can estimate the dielectric constant of liquid substances without the experimental limitation.

The miniaturization of high-power laser system, especially the space-based laser system, poses a great threat to the national space security. In this paper, facing the space-based offensive and defensive operations in the future, the features and damaging modes of high-power laser system as well as the conventional laser protection technologies of spacecraft are analyzed. On this basis, a novel multilayer protective armor is proposed, which can significantly improve the survivability of spacecraft under the high-power laser beam. Furthermore, a physical-informed deep surrogate model for the temperature field reconstruction is also studied to achieve the laser spots detection within second, helping the on-orbit spacecraft to implement the evasive maneuver. The results of this work can help enhance national space safety level.

Infrared dim and small target detection plays an important role in different applications, and many infrared dim and small target detection methods have been proposed. However, most existing methods are based on single camera. Although they have achieved good performance, most of them have poor performance in complex scenes with dim and small targets. Based on this, we introduce the array camera, which can provide more view information, so as to obtain better performance in the complex scene of dim and small targets. Furthermore, in order to further improve the confidence of detection rate, we introduce a probability estimation module into our method. Specially, the array camera provides more target position information through the view-correlation frames. And the probability estimation module is introduced to fuse the information of each view. Extensive experiments on different scenes demonstrate that our method achieves better performance in dim and small target detection, and obtains higher confidence of detection rate and lower false alarm.

In the field of electronic modulation, vanadium dioxide (VO₂) is a potential material owing to its function of automatic insulator-to-metal transition (MIT) which can induce rapid changes in electrical resistivity through MIT. Nevertheless, the application of modulator based on VO₂ is limited by some performance shortcomings, including wide hysteresis loop width (ΔH), high phase transition temperature (T_c) as well as low phase transition amplitude (A_MIT). In this work, by DC (DirectCurrent)-magnetron sputtering with doping Fe³⁺ into VO₂ films, narrowed ΔH and decreased Tc are observed. Interestingly, the Fe doped VO₂ films show ultra-high phase transition amplitude despite the low T_c due to the influence of Fe dopants. Specifically, the 0.5% Fe-doped VO₂ film shows the best MIT characteristics with ultra-high phase transition amplitude of 10⁴, narrow ΔH decreased to 9.8° and low T_c around 60.02°C, which is considered to be the first time to highly heighten the electrical MIT properties by Fe doping. In addition, we also comprehensively studied the influences of doping with Fe element on the MIT properties and microstructures based on characterization such as SEM, XRD, Raman shift and XPS results. These results show that our unique preparation method can manufacture VO₂ thin films with excellent MIT properties, which will be beneficial to the popularization and publicity of VO₂ based electrical modulator.

The lunar is a stable radiation source, which can be used as an ideal source for in-orbit calibration of remote sensors and evaluation of detector degradation performance. A passive lunar calibration method is proposed for Chinese ocean color satellite, which reuses field of view of cryogenic-deep-space calibration, periodically achieves monthly lunar calibration tasks. This work enriches ocean color satellite in-orbit calibration methods and improves data accuracy of products. The start and stop angle vector of cryogenic-deep-space, satellite-lunar pointing vector and imaging observation model are established in simulation software. The satellite and payload parameters are used as input conditions to carry out the simulation of the lunar calibration timing. The lunar calibration timing of payload COCTS (Chinese Ocean Color and Temperature Scanner) is simulated 00:00:00~24:00:00UTC on June 28, 2020. The result shows that lunar calibration was carried out for twice. The starting UTC (Universal Time Coordinated) time was 15:16:44 and 16:56:37, respectively. The duration was two seconds. Through analyzing the 0-level products of in-orbit satellite received by the application system, the cryogenic-deep-space data showed abnormal changes at 15:16:45 and 16:56:38 on June 28, 2020, and the DN (Digital Number) values dropped from 300 to 62 and 73, respectively. The in-orbit calibration timing is basically the same as the simulation results, and the numerical anomalies of cryogenic-deep-space data are consistent with the principle design, indicating that the simulation model can be used to predict the in-orbit lunar calibration timing of ocean color satellite. The analysis method can be used for cryogenic-deep-space imaging mission mode and payload design of the follow-up ocean color satellites.

Single-photon Lidar (SPL) has many applications, ranging from geographic surveying to autonomous navigation. For accurate range extraction and range-walk-error (RWE) correction, a good TCSPC histogram is usually required and collected by using thousands of accumulations. While in high-flux regime, using a few accumulations is possible to perform range extraction and RWE correction and could speed up data acquisition process. This paper proposed a method to deal with two challenges of depth estimation when only a few accumulations (less than 10 times) are applied, including the extreme sparsity of photon-counting data and the non-steady state in free-running SPL. We first developed a forward model for non-steady state free-running SPL. As a second step, we computed the initial depth through our coarse depth estimation method, which incorporated the maximum of cross-correlation, threshold filtering, and cluster analysis of potential range candidates. Then solving an inverse problem through our forward model and initial values gives the estimate of the depth. Monte Carlo simulated results indicated good improvement in performance of detection probabilities and RWE correction. As compared with a baseline method, our method maintains a same detection probability while shows a 26% less false alarm probability on average. When signal-background ratio SBR is 0.2 (S = 2, B = 10) and accumulating times is only 4, our method gives a 75% less RWE. Our method could enable improvement of depth estimation from sparse photon-counting data collected by a few accumulations in SPL.

Light field (LF) imaging has proven to be a promising technique in computer vision field. However, there is a tradeoff between spatial and angular resolution of LF images, which limits the application of LF cameras. Super-resolution (SR) of the angular domain is proposed to improve the angular resolution of LF images. However, most of the SR frameworks cannot adapt to LF datasets with multi-size disparities, especially the large disparities. In this paper, we proposed a learning-based SR framework named EASRnet. The EASRnet consists of three parts - Disparity adaptation, Feature extraction, and Feature restoration parts, and achieves angular SR tasks by using residual blocks and a structure with branches to reconstruct high-frequency details of up-sampled epipolar plane images (EPI). It employs an additional blur layer to accommodate LF datasets with different disparities. The experimental results show that the proposed approach can reconstruct novel view images with satisfactory accuracy.

In this paper, an effective and simple internal model control (IMC) with extended Kalman filter (EKF) is presented for low-velocity smoothness control of air bearing linear feed stage (ABLFS) driven by permanent magnet synchronous linear motors (PMLSMs). Firstly, the ABLFS is modeled as an inertia system with the nonlinear force ripples which correlates with phase current. The identification experiment is conducted using white noise method to get the approximate linear model of the air-bearing stage. Then, a typical linear feedback controller is derived from the standard IMC principle for the fundamental close-loop control. However, the IMC controller is sensitive to disturbances and modal uncertainties which limits the tracking accuracy. To overcome these drawbacks, the Kalman filter is employed as a state observer which makes the optimal estimation of system state variables (current, velocity and displacement). Since the Kalman filter is a synthesis process by calculating the probability density of measured values and predicted values, the results are more reliable to take place of the encoder feedback. Finally, the effectiveness of this modified method is validated by comparative experiments on a practical ABLFS system.

Aiming at the influence of the atmosphere on the measurement accuracy of ground-based interferometric radar (GB-InSAR) in landslide monitoring, the article discusses four algorithms of conventional linear compensation, meteorological data correction, distance function fitting, and nonlinear phase compensation. And proposed an optimization scheme based on the fixed-point method of distance function fitting algorithm. This scheme selects the stable PS point through the double threshold of the cross-correlation coefficient and the phase standard deviation, and then performs the least square fitting on the atmospheric disturbance distance function model, so as to achieve high-precision atmospheric disturbance phase compensation. Finally, the performance of the four atmospheric compensation algorithms is simulated and analyzed using real data. The results show that the compensation algorithm is effective and greatly improves the accuracy of the phase measurement error.

In the assembly of space camera, the assembly of large aperture mirror is the key of the system assembly. In this paper, the flexible supported mirror of space camera is taken as the research object, and a process route is proposed to complete the micro-stress assembly of the mirror and the transmission from optical datum to mechanical datum at one time by taking the adhesive assembly of the mirror and the transmission from optical datum to mechanical datum into consideration. According to this process route, the assembly process design and installation platform construction are completed. In addition, the factors affecting the micro-stress assembly of the mirror are analyzed theoretically and experimentally, and the transfer accuracy of the optical-mechanical reference is analyzed. This process method has been successfully applied to the assembly of a space camera's primary mirror. After the assembly, the shape of the primary mirror RMS is less than λ/50, and the Angle between the optical axis and the normal of the structure's mounting surface is less than 10". The process method presented in this paper provides a technical reference for the assembly of similar optical-mechanical systems.

Electromagnetic wave reverse time migration (RTM) imaging based on finite difference time domain (FDTD) method is widely used in geological exploration. It can accurately attribute the wavefield to its true subsurface spatial location when dealing with steeply dipping structures and complex velocity models. However, the forward and backward propagation processes of RTM imaging require recording wavefield information at each time step, which causes a large number of buffer accesses to degrade the overall calculation speed. This computationally and data intensive features limits its application in more scenarios. Therefore, this paper develops the FDTD prototype accelerator based on FPGA to accelerate the forward and backward propagation process of reverse time migration imaging, and realize the hardware circuit of cross-correlation imaging conditions. The experimental results show that the forward propagation process of FPGA at a clock frequency of 100MHz is 17.384 times faster than the PC using 3.6GHz CPU, and the backward propagation process is 18.371 times faster in comparison.

Aiming at ascent trajectory planning of pre-cooled combined power vehicles, in order to ensure the optimal flight performance of pre-cooled combined power engines, the paper conducts a trajectory planning study on a pre-cooled combined power vehicle from the end of take-off to the first and second separation point ascent, considering the flight mission and the pre-cooled combined power engines’ characteristics of different operating modes. A nominal trajectory is planned based on the height-velocity profile, and then the nominal trajectory is simulated and verified by linear feedback control methods.

In aquaculture, the normal growth of fish is closely related to the density of aquaculture. Therefore, it is of great significance to use remote sensing images to accurately segment the cages in a specific sea area at a macro level. This research proposes an accurate segmentation scheme for remote sensing cages based on U-Net and voting mechanism. Firstly, a remote sensing cage segmentation (RSCS) data set is produced, which includes fifty-three high-resolution cage images with inconsistent resolution. Secondly, by using random cropping and data enhancement operations on the training samples, three training sets with image block sizes of 256×256 pixels, 512×512 pixels, and 1024×1024 pixels are created. And through the introduction of U-Net network, three training sets of different sample sizes are trained separately and three trained models are generated. Then, after reasonably filling the test image, a window sliding overlap cropping method is adopted. The high-resolution remote sensing cage test images are sequentially cut into the image blocks for segmentation, and the segmented image blocks are spliced and combined into the binary segmentation image by the mean method. Finally, for each image, the three binary segmentation images generated by different trained models are used to vote for each pixel. The experimental results show that by testing three remote sensing images of Li'an Port, Xincun Port and Potou Port, the Mean Intersection over Union (mIoU) is 0.865. Our data and code can be available online.