2021 Vol. 42, No. 1
Aiming at the test results of minimum resolvable temperature difference (MRTD) of a multi-band common aperture long focal length reflective infrared imaging system deviating from design value under laboratory conditions, starting with the definition of MRTD, the influences of observation frequency response, visual angle, sampling phase transfer function and other factors affecting MRTD test results were analyzed in detail. Secondly, the influence of sampling phase on infrared system static and dynamic MRTD test results was emphasized. It was demonstrated that static MRTD test value would deviate from the design value when spatial frequency was in the range of
As an important evaluating method, the technology readiness level (TRL) can provide a better decision-making basis for the defense acquisition of weaponry, and can effectively reduce the technical risk of equipment development. The foreign evaluation methods of technology readiness level were briefly introduced, and the application of technology readiness level in the American ship-based laser system LaWS was studied. The LaWS system consists of five sub-systems such as transmitter-telescope, light source in military operation, diastimeter, target tracking sensor and control system. In the system development, in addition to a small number of custom components, a large number of industrial lasers, inertial measurement devices, sensors, video trackers and other commercial devices are adopted. A series of demonstration experiments were retrospected in LaWS system, the technical problems exposed in each test and the corresponding solutions adopted were analyzed, and the technology readiness level of each stage was evaluated. Finally, the research and development approach and the significance of battlefield deployment were reviewed.
In order to meet the requirements of electro-optic payload on vibration isolation performance, a high static stiffness and low dynamic stiffness nonlinear vibration isolator with rhombus linkage mechanism as negative stiffness component (rhombus HSLDS vibration isolator) was proposed. The mathematical model of vibration isolator was established by statics analysis method, and stiffness parameter setting and nonlinear adjustment methods were studied; the harmonic balance method (HBM) was used to solve the kinetic equation, and influence of parameters on the vibration isolation performance was analyzed; the theoretical model and conclusions were verified by kinetic simulation software ADAMS and prototype. The test results show that the rhombus HSLDS vibration isolator has convenient parameter adjustment capability, and the zero-position stiffness and stiffness nonlinearity can be set and optimized by parameters of tension spring and linkage. And the optimization effect of vibration isolation caused by stiffness nonlinearity is affected by the damping of main vibration isolator and zero-position stiffness parameter. Compared with traditional linear vibration isolator, the rhombus HSLDS vibration isolator has significant advantage of nonlinear vibration isolation and can better meet the needs of electro-optic payload vibration isolation requirements.
In order to improve the installation and adjustment efficiency of off-axis three-mirror optical antenna of space laser communication, it is necessary to optimize the design of total freedom of the mirror. Based on the principle of coaxial three-mirror afocal system, a method of integration of primary-tertiary mirror was proposed. The relationship between structural parameters of optical system was deduced, and an off-axis three-mirror optical antenna with compact structure and primary-tertiary mirror integration was designed by Zemax optical software. The design results show that the optical image quality of full field of view is better than diffraction limit, the primary mirror and tertiary mirror are close to each other in space, the common mother plate can be processed and surface shape can be detected by the common reference, which provides a method for integrated processing of primary mirror and three-mirror. The degree of installation and adjustment freedom of optical antenna is reduced by 6, which reduces the difficulty and improves the efficiency.
Many applications expect to search for targets with a large field of view (FOV) , and identify the targets in a relative small FOV. A miniaturized two-step 7.5× zoom system was proposed and designed, which worked in the visible band of 486 nm~656 nm with a wide-angle FOV of ±15°, telephoto FOV of ±2°, F-number in both FOV of 2.8, and overall length of 60 mm. The system consisted of glass lenses and plastic-injection lenses, which contained one front fixed group and one movable group. The conversion of wide-angle search function and long-focus gaze function was completed by switching the movable group between two zoom positions, which had the advantages of miniaturization, lightweight and low cost.
The structural design and performance test of the sub-mirror co-phase correction device were finished. The device covered two-stage structure and was mainly composed of coarse adjustment stage, fine adjustment stage and lateral unloading mechanism. Based on the finite element method, the overall design and the performance analysis of the device were finished, and then the prototype development and basic performance test were also completed. Finally, the test results indicate that the displacement of adjustment system is about ±2.5 mm, the accuracy is about 30 nm RMS, and the natural frequency is about 70.3 Hz with analog mirror, which are in good agreement with the finite element analysis results. In conclusion, the performance of the design meets the requirements of the segmented-mirrors.
In advanced 3D dose verification system for radiotherapy based on scintillation light field imaging, it is necessary to use the point spread function to extract the real scintillation light data of each 2D plane from the overlapping images. The point spread functions of the light field camera at different refocusing positions are measured by knife-edge method combining the digital refocusing of light field and the focussing ranging method. And when the refocusing plane refocused at α=0.6, the function expression of Gaussian defocus for checkerboard calibration plate at α=0.7 was given. The law of light field camera point spread function was studied. All point spread functions could be obtained through function fitting by measuring five necessary values and the workload of point spread function calibration caused by the increase of the number of layers in optical layered imaging was reduced by using this law. Introducing the research results into image processing could obtain the real scintillation data of each 2D refocusing surface through deconvolution operation, which would be helpful for the real-time and accurate measurement of 3D dose.
The depth estimation of objects in the scene is a key issue in the field of the unmanned driving. The infrared images are helpful to solve the depth estimation problem under poor light conditions. Aiming at characteristics of unclear infrared images texture and insufficient edge information, a combination of attention mechanism and graph convolutional neural network was proposed to solve the problem of monocular infrared images depth estimation. First of all, in the depth estimation problem, the depth information of each pixel in the image was not only related to the depth information of its surrounding pixels, but also needed to consider the depth information of a larger range of other pixels. The attention mechanism could be effectively extract the pixel-level global depth information association of images. Secondly, the features obtained based on the depth information association could be considered as non-Euclidean data, and the graph convolutional neural network (GCN) was further used for reasoning. Finally, in the training phase, the continuous depth estimation regression problem was transformed into the classification problem, which made the training process more stable and reduced the learning difficulty of the network. The experimental results show that the proposed method has obtained good results on the infrared data set NUST-SR. When the threshold index is less than 1.253, the accuracy rate is improved by 1.2%, which is more advantageous than other methods.
Semantic information is essential for mobile robots to understand the content of the environment and perform complex tasks. Aiming at the problem that the point clouds constructed by ORB-SLAM2 is too sparse and lacks semantic information, a dense point cloud semantic map of the environment by combining the object detection algorithm with visual SLAM technology was constructed. First of all, the object detection network YOLO v3 and object regularization were used to accurately obtain the 2D label of the object. At the same time, the ORB-SLAM2 algorithm was used to construct the environment's sparse point cloud map. The color image with 2D labels, corresponding depth images, and key frames were used to generate dense point cloud labels with semantic information. Then the graph-based segmentation algorithm was used to segment the dense point cloud, and the point cloud labels were fused with the segmented point cloud so as to construct a dense point cloud semantic map of the environment. The proposed method was tested on the TUM public data set and the experimental results show that the method can construct a better semantic map. Compared with the traditional ORB-SLAM2 algorithm, this system reduces the absolute pose error and absolute trajectory error of the camera by 16.02% and 15.86% respectively, in the process of constructing the map, which improves the mapping accuracy. In order to reduce the storage space of point cloud maps and facilitate mobile robots' navigation and avoidance, the constructed semantic maps are finally converted into octree maps.
In order to provide the accurate initial physical information for numerical simulation of aircraft and missile separation, a missile physical information extraction scheme based on regional Hough transformation was proposed, which used high-speed cameras to extract the positions of marking circle centers on the missile body and tail in the video frame, and calculated the physical information of missile through the coordinate transformation. In frame processing, according to short time interval and negligible change of velocity vector in the frames, a circle search algorithm based on regional Hough transformation was proposed. In view of the fact that the real motion vector is biased rather than uniformly distributed in search window, the search box was fixed shifted. The centering error between predicted region and actual region is less than 5%, the consuming time is reduced by more than 41.6% compared with the traditional algorithm, and the average error between calculated results and actual results does not exceed 2.39%. Also, this algorithm can adapt to the complex backgrounds and meet the timeliness and accuracy requirements of the program.
In order to obtain the high quality all-in-focus imaging in large depth of view, an all-in-focus image reconstruction technology based on single exposure light field imaging and guided filter was proposed. First, the field of view information was collected by the optical field imaging, and the multi-focus image source set was obtained by the optical field reconstruction. Then the guided filter method was used to determine the weight of image fusion at all levels, and finally the image fusion was performed to obtain the all-in-focus image with extremely large depth of field. The experimental results show that the proposed method not only effectively ensures the background consistency, but also keeps the edge retention to reconstruct the all-in-focus image with high quality. Therefore, the proposed technology can be potentially adopted in many applications such as on-site monitoring, geographical exploration, military reconnaissance and unmanned technologies, etc.
In the process of applying robots to car body spot welding quality inspection, the positioning accuracy of its welding spots was affected by factors such as quality of spot welding operations and car body manufacturing errors, which resulted in actual welding spots did not coincide with designed values. Aiming at the problem that traditional teaching could not perform real-time compensation for welding spot positioning, a welding spots positioning strategy based on binocular vision guided robots was proposed, and a support vector machine regression error compensation model optimized based on improved particle swarm algorithm was constructed to compensate the positioning results. The binocular sensors were installed at the end of the robot, the binocular positioning principle was used to initially locate welding spots, and the measured data and actual data of welding spots position was used as learning samples. The trained error compensation model was used to predict positioning error of the system, and the compensation results were used as the correction value guiding the robot to locate welding spots. The experimental results show that the positioning accuracy after compensation is greatly improved, which verifies the effectiveness of the method.
In order to achieve accurate extraction of the center of laser stripes during the underwater detection of nuclear fuel rods, an self-adaptive stripes center extraction method for reflective surface of underwater nuclear fuel rods was proposed. According to characteristics of water scattering and object surface high reflections in detection environment, the underwater noise points and reflective noise points were removed to realize segmentation and extraction of laser stripes; the curve fitting of BP neural network and adaptive convolution template generated from light bar geometry information were utilized to realize contour and gray distribution correction of reflective region, so that the gray distribution of light bar section conformed to the Gaussian distribution; the subpixel precision location and extraction of laser stripes center were realized in light bar section direction by gray centroid method. The experimental results show that this method can effectively solve the problems of discontinuous center line and many noise points of reflective surface light bar of measured object. The 3D reconstruction error of point cloud is within 0.2 mm, which ensures accuracy and stability of stripes center extraction and meets the engineering requirements of underwater detection of nuclear fuel rods.
In order to ensure the quality of whole optical system, it is important to measure and check the optical curvature radius accurately. The method of mechanical spherometer and the method of optical projection were combined, and the photoelectric image method was used to measure the vector height, then the radius could be calculated indirectly. The influence caused by edge error of measured element on the accuracy of the vector height and curvature radius was analyzed. Polarization imaging and general imaging were used to measure the height and compare the results, and it is found that polarization images can have better edge details, the accuracy of the vector height was effectively improved, and the error of the optical curvature radius was reduced by more than 0.5%. The results show that the method of polarization imaging has important application value in measuring the optical curvature radius.
The absolute test technology of optical surface shape avoids the limitation of reference surface shape accuracy of interferometer and can effectively improve the detection accuracy of flat-crystal surface shape. The absolute test technology of two-flat crystal three-sided mutual test of N-position image rotation method was adopted to solve the three-dimensional absolute surface shape distribution of tested flat-crystal, and the results included the information in medium frequency band. The recursion formula was used to construct the N-position virtual rotation results of rotational change item. After summing and averaging, the rotational change item was obtained, and the tested wave surface shape was obtained after superposition of the rotational invariant item. The theoretical error of the algorithm was derived, the rotation angle was optimized and the number of virtual rotation was increased, and the accuracy of the algorithm was improved. After optimization, the simulation results show that the RMS value accuracy of residual wave surface is 0.14 nm. The two-flat crystal three-sided mutual test of flat-crystal with 150 mm caliber was carried out, and the experimental results were compared with those of the traditional three-sided mutual test method. The RMS value deviation is less than 0.5 nm, which verifies the accuracy of the algorithm.
A 3D measurement system based on the combination of image bundles and binocular structured light was proposed. It could measure the three-dimensional shape of small objects under the environment of detonation and impact, which could not be directly measured. The structured light was directly projected onto the surface of the measured object and the distorted fringe images were imaged to the front side of the image bundles. Then those images were collected into the camera by another imaging lens after transmitted to the other side of the image bundles. The height of the measured object could be reconstructed by the binocular matching as long as the corresponding phase distributions were restored by the three-frequency phase unwrapping algorithm. The experimental results show that the measuring system via image bundles can fulfill the 3D morphology reconstruction of the small objects with high precision and can bring the advantages of flexibility due to the introduction of image bundles.
In the fringe projection profilometry system based on triangulation principle, the fringe projected on the reference plane always produces the period broaden phenomenon, which can cause phase distortion and even affect the measuring accuracy. The fringe position was taken as the control variable to derive the linear mathematical model of fringe period correction. The model parameters were obtained with a simple and convenient calibration process. Then the new projected fringe was calculated according to the correction model, and the periodic distribution fringe was obtained on the reference plane. The experimental results show that the varying range of the fringe period after correction is within ±0.1 pixel, and this method can obtain the more accurate 3D profile measurement results.
Aiming at the problem of vertical target coordinate measurement when two projectors were touching the target at the same time, a two-target identification method of circular array photoelectric detection system was proposed. A circular detection array was formed by using photodetectors, and three fan-shaped linear lasers with all luminescence angles of 60° were uniformly arranged on circular detection array to form a detection light curtain. When the two projectiles passed through detection light screen at the same time, six projectiles would be generated on the circular detection array. The central position of the six projectiles would be identified through the signal processing circuit. Finally, the target coordinates of two projectiles would be calculated by system projectile hitting coordinate measurement formula. On the basis of the theory of system measurement, the coordinate measurement model of the projectile hitting target was established, and the coordinate measurement error was analyzed and simulated. The simulation results show that the maximum standard deviation of the X coordinate measurement error is 2.7 mm and the maximum standard deviation of the Y coordinate measurement error is 0.6 mm when the target surface is 1 m×1 m. The experimental results show that the standard deviation of X coordinate measurement error is 2.22 mm and Y coordinate measurement error is 1.98 mm when the target surface is 1 m×1 m. Therefore, the system proposed can effectively measure the target coordinates of double projectors with a projectile diameter of 4.5 mm or above.
At present, machine vision technology is widely used in the field of impact point coordinate measurement. In order to solve the problem of binocular vision technology calibration and complex calculation, a method of impact point measurement based on monocular vision technology was proposed. The four corner points of the target surface and the pixel coordinates of the impact point was located through image processing technology, the rectangular P4P method was used to calculate the posture of the target surface coordinates relative to the camera coordinate system, and the actual coordinates of the impact point on the target surface was calculated according to the principle of monocular vision imaging. According to this method, the coordinates of target bullet hole in the experiment were successfully measured, and the maximum test error was 2.3 mm. The results show that the method can quickly and accurately measure the position and attitude relationship between the target surface and the camera and the coordinates of the impact point.
Electron bombarded (EB) CMOS sensor process defects cause blind elements in the image, which has a greater impact on image quality. A new blind element detection algorithm based on cross calculation of multiple images was proposed. For untargeted images with different illumination brightness, by setting different thresholds, the image was binarized; then multiple binarized images were calculated according to the position of “bright spots”, and the blind element marking template image was generated. Finally, classification and selection were performed according to the position of blind element, and the final blind element mark template was obtained. The final blind element mark template information provided a reliable basis for evaluating the quality of EBCMOS sensor and the blind element compensation in actual follow-up products.
The ring laser gyro (RLG) light intensity mode-scanning curve is the basis of the frequency stabilization, and also is the basic characterization of the oscillation characteristics for RLG ring laser and is one of the important signs of the gyro performance. By introducing a new parameter which called mode height, other characteristics of the light intensity mode-scanning curve were described. The mode height of the light intensity mode-scanning curve was the maximum variation of the corresponded light intensity under conditions of one 0.632 8 μm cavity length change, which was a characteristic quantity reflecting the sharpness of the light intensity mode-scanning curve. This characteristic quantity was directly related to the gyro frequency stabilization accuracy and the frequency stabilization response time. The higher the mode height, the higher the corresponding frequency stabilization accuracy, and the shorter the frequency stabilization response time. The physical mechanism of the mode height in RLG scanning process was analyzed theoretically, and the main factors affecting the mode height were determined. Through the theoretical analysis and the numerical simulation, it is concluded that: by increasing the curvature radius of spherical mirror, increasing the cavity length, reducing the loss and increasing the gain, the mode height of the RLG light intensity mode-scanning curve can be increased more than 3 times, which has important guiding significance and practical engineering value for improving the frequency stabilization accuracy, shortening the frequency stabilization response time, reducing the nonlinear error of gyro proportional factor, and improving the fast stability of RLG.
In the process of the laser ultrasonic surface defects detection, the quantitative characterization of the defects mainly depends on the operator's judgment, and it is easily interfered by the human factors, which leads to the unstable detection results. To solve this problem, an defects automatic classification detection method based on the two-dimensional convolutional neural network (2D-CNN) for image recognition was proposed. The finite element method was used to simulate the laser ultrasonic detection process, and the ultrasonic signal data was collected for training the classification model; the continuous wavelet transformation (CWT) was used to process the ultrasonic signal to obtain the wavelet time-frequency images, and the images were used as inputs to train the convolutional neural network (CNN) classification model to realize the automatic classification of the surface defects depth. The verification results show that the proposed detection method can accurately classify the defects of different depths, and the average accuracy rate of the test reaches 97.3%; the constructed CNN classification model can independently learn the defects features of the input images and complete the classification, which improves the stability of the test results, and provides a new idea for the automatic analysis and processing of laser ultrasonic defects detection.
With the advantages of high resolution, high sensitivity, wide spectral coverage and rapid measurement, the dual-comb spectroscopy technology has developed rapidly in the field of molecular and atomic spectroscopy. According to the principle of dual-comb spectroscopy and taken the number of pump light sources used as the classification criteria, the principles of multiple dual-comb spectroscopy techniques are briefly introduced basen on different implementation schemes at home and abroad in recent years, the various experimental schemes were described proposed by different experimental groups with different numbers of pump light sources, and the results of their spectral experiments were introduced. With the deepening of research, reducing the number of pump light sources used is one of the important directions for the development of dual-comb spectrometers to achieve miniaturization of equipment, and the increase is an important development idea from one-dimensional to multi-dimensional. The classic dual light source configuration has made important progress in accuracy and two-dimensionalization.
The laser four-quadrant detection system is an important part of laser guided weapons. The accurate evaluation of its performance in the design stage is conducive to the iteration and optimization of the design scheme. The simulation model of the four-quadrant detection system was established based on LitghtTools. Through the establishment of target model, optical system model and four-quadrant detector model, the system could be effectively simulated. Combined with principle of four-quadrant detection system, the sum-difference amplitude of angle deviation could be obtained; the stray light of the system could also be simulated to evaluate the impact of stray light on system; in addition, it could also simulate the laser speckle effect and analyze the influence of speckle characteristics on the sum-difference amplitude. Therefore, the modeling and simulation based on LitghtTools provided an accurate basis for design optimization. Finally, the simulation model was verified by experiments. The sum-difference amplitude of the simulation is basically the same as that of the experiment in the range of ± 15°.
LED has become a mainstream light source in helmet liquid crystal (LC) display. To increase luminance and decrease backlight unit volume, a flatening design of double free-form surface lens was carried out to form the lens film. The simulation results show that the central luminance increases and the spot size decreases after lights passing through the lens film. The feedback optimization was performed to redefine the power distribution when designing the double free-form surface lens, and the lens film is redesigned. The simulation results illustrate that the spot size increases from 2.3 mm to 7.1 mm after optimization. The designed lens films were prepared by maskless direct-writing lithography, and each film contained a free-form surface. The backlight units with and without the lens films were measured, and the measured results show that the average luminance with the lens film increases by 18.1%. The designed lens film can effectively increase the luminance and decrease the backlight units volume.
In order to improve the mechanical properties of K9 optical glass, which used in some special fields such as heavy pressure and drastic temperature changing, etc, and ensure the optical properties can meet the requirements of precision optical instruments, the chemical strengthening technology for K9 optical glasses was studied. Based on the theory of crack propagation during the fracture process of brittle materials, the calculation model of stress intensity factor for chemical strengthened K9 optical glasses was formulated. The relations between glass surface stress, surface micro-crack depth and toughness were discussed, and several problems in the chemical strengthening process needed to attention were pointed out. By experimental research, the effects of strengthening temperature and time on the strength, surface stress and stress depth of K9 optical glasses were analyzed. The chemical strengthening process for K9 optical glass was optimized, the strengthening temperature is 400 ℃ and the strengthening time is 40 h. The measured results of mechanics and optical properties for K9 optical glasses with size of 220 mm×110 mm×22 mm are obtained. The measured results show that the surface stress is 500 MPa, the stress depth is about 50 μm, the bending strength is increasing more than 3.5 times, and the optical discrimination and transmittance show no obvious changing.
A multi-wavelength thulium-doped actively mode-locked fiber laser was introduced, in which the gain medium was a 2 m thulium-doped fiber, and the active mode-locking could be realized by LiNbO3 intensity modulator. An optical filter based on birefringence was added into cavity, the birefringence filter effect of polarization maintaining fiber was used to filter out the superfluous supermode noise in cavity, and also the multi-wavelength output could be realized. The mode-locked pulse frequency spectrum signal-to-noise ratio could reach to 68.48 dB at fundamental frequency, and the maximum number of wavelength channels was 5 in a stable mode-locked state. Furthermore, the polarization independent isolator in cavity was replaced by polarization dependent isolator, and mode-locked pulse was modulated by digital signal. The optical signal-to-noise ratio of eye diagram can be increased by 8.67 dB, which indicates that time stability of mode-locked pulse can be improved effectively.
The traditional fiber grating sensors are cross-sensitive to temperature stress and cannot simultaneously measure the change of the temperature stress of the measured object. In view of this situation, a new method of lithographing phase-shifted gratings using ultraviolet light was proposed. Before lithographing the gratings, the method of electrode discharge was used to remove the photosensitivity of a very small section of the optical fiber, so that the original uniform periodic distribution of the fiber was destroyed to form a phase-shifted grating, and its theoretical analysis was carried out. There were two distinct resonance peaks in the transmission spectrum of this phase-shifted grating. Using the properties of two peaks with different sensitivity to temperature and dependent variable, the simultaneous measurement of temperature and stress could be achieved by establishing the demodulation matrix. The experimental results show that the phase-shifted Bragg grating made by this method can realize the simultaneous measurement of temperature and stress accurately. The maximum temperature sensitivity of the sensor can reach to 9.51 pm/℃, and the sensitivity variance is lower than 2.125×10−7. The maximum strain sensitivity can reach to 0.767 pm/με, and the sensitivity variance is lower than 2.156×10−10.