Advance Search
中文

2022 Vol. 44, No. 1

column
  Previous Issue | Next Issue 
Display Method:
Image Processing and Simulation
Infrared and Visible Image Fusion Algorithm Based on the Decomposition of Robust Principal Component Analysis and Latent Low Rank Representation
DING Jian, GAO Qingwei, LU Yixiang, SUN Dong
2022, 44(1): 1-8.
Abstract HTML (86) PDF(60)
Abstract:
The fusion of infrared and visible images plays an important role in video surveillance, target tracking, etc. To obtain better fusion results for images, this study proposes a novel method combining deep learning and image decomposition based on a robust low-rank representation. First, robust principal component analysis is used to denoise the training set images. Next, rapid latent low rank representation is used to learn a sparse matrix to extract salient features and decompose the source images into low-frequency and high-frequency images. The low-frequency components are then fused using an adaptive weighting strategy, and the high-frequency components are fused by a VGG-19 network. Finally, the new low-frequency image is superimposed with the new high-frequency image to obtain a fused image. Experimental results demonstrate that this method has advantages in terms of both the subjective and objective evaluation of image fusion.
Hyperspectral Image Hybrid Convolution Classification under Multi-Feature Fusion
XIONG Yu, SHAN Deming, YAO Yu, ZHANG Yu
2022, 44(1): 9-20.
Abstract HTML (47) PDF(39)
Abstract:
To address the problem of insufficient utilization of spatial-spectrum features in existing convolutional neural network classification algorithms for hyperspectral remote sensing images, we propose a hyperspectral image classification strategy based on a hybrid convolution capsule network under multi-feature fusion. First, a combination of principal component analysis and non-negative matrix decomposition is used to reduce the dimensionality of a hyperspectral dataset. Second, the principal components obtained through dimensionality reduction are used to generate a multidimensional feature set through super-pixel segmentation and cosine clustering. Finally, the superimposed feature set is used to extract spatial-spectrum features through a two-dimensional and three-dimensional multi-scale hybrid convolutional network, and a capsule network is used to classify them. We performed experiments on different hyperspectral datasets, and the results revealed that under the same 20-dimensional spectral setting, the proposed strategy significantly improves the overall accuracy, average accuracy, and Kappa coefficient compared to traditional classification strategies.
Infrared Image Non-uniformity Correction Algorithm Based on Full Convolutional Network
MOU Xingang, CUI Jian, ZHOU Xiao
2022, 44(1): 21-27.
Abstract HTML (82) PDF(79)
Abstract:
The infrared imaging system will still exhibit non-uniform noise after two-point correction. An infrared image non-uniformity correction algorithm based on a fully convolutional deep learning network was proposed in response to this problem. This algorithm combines the subnetwork and main network for non-uniformity correction. The network contains a nonuniformity-level estimation subnetwork. After inputting the infrared image with non-uniform noise into the non-uniformity level estimation subnetwork, the outputted non-uniformity level estimation map is input into the main network together with the original noise infrared image. The non-uniformity level estimated map generated by the subnetwork prevents the network from overfitting only for the non-uniformity of the same grade. After experimental verification, the algorithm overcomes the problem of edge blur generated by the scene-based algorithm. The algorithm will not appear blurred, the images have high definition and rich details, and the quality of images is good.
Infrared Structured Light for 3D Face Reconstruction
TANG Shiyang, ZHU Jiangping, ZHANG Jianwei
2022, 44(1): 28-32.
Abstract HTML (17) PDF(65)
Abstract:
In structured-light 3D face reconstruction, it is easy to lose detailed data and obtain reduced modeling accuracy, which leads to low integrity and poor recognition of 3D faces. In this study, a binocular three-dimensional camera system based on an infrared fringe is developed. The wrapped phase is generated using the phase-shift method by projecting infrared fringe-structured light. The absolute phase is obtained using the three-frequency method, and a parallax diagram is generated to obtain a three-dimensional face model. Experiments reveal that measurement errors for the distance of sphere centers are less than 0.1% when measuring standard spheres, and face accuracy is within 0.1 mm. For the eyes, eyebrows, and other areas with weak texture, data loss is reduced, which is better than visible light. The face model varies more smoothly and is more consistent with the real face. This preliminary analysis of the performances of the two proposed techniques can be used as a reference for further comparisons in the analysis of various techniques and algorithms.
Fusion Reconstruction Method for 3D Temperature Fields on the Human Body Surface
YANG Yanlong, XU Chao
2022, 44(1): 33-40.
Abstract HTML (38) PDF(54)
Abstract:
Reconstruction of 3D temperature fields on the human body surface can provide reliable data for a number of human medical analyses, including diagnoses. Based on the limitations of infrared imaging, such as poor temperature measurement accuracy, insufficient imaging resolution, and poor display effects, the reliability of the 3D temperature field collected using infrared imaging is low. To overcome these problems, we propose a fusion reconstruction method for 3D temperature fields on the human body surface. First, the blackbody temperature measurement and calibration method is used to correct the errors in the temperature measurement results of an infrared thermal imager. Second, contrast enhancement processing is applied. Third, super-resolution processing is used to make the infrared images match the 3D data in terms of spatial resolution. Finally, in the data fusion stage, based on the fact that the target feature points extracted from different images correspond to the same position in the space, the system structure parameters obtained through calibration are corrected. Experimental results demonstrate that the temperature error of the 3D temperature field is less than 0.26℃, the 3D distribution of the temperature field is improved, and the display effect is enhanced.
Image Deblurring Method Based on a Dual-Discriminator Weighted Generative Adversarial Network
HUANG Mengtao, GAO Na, LIU Bao
2022, 44(1): 41-46.
Abstract HTML (23) PDF(27)
Abstract:
The original generative adversarial network (GAN) is susceptible to the problems of vanishing gradients and mode collapse during the training process, and its deblurring effectiveness is poor. This study proposes an image deblurring method using a dual-discriminator weighted GAN. To extend the original GAN, a discriminator network is added to combine the forward and reverse Kullback–Leibler (KL) divergences into an objective function, and weights are used to adjust the ratio of forward and reverse KL divergences to leverage the complementary characteristics of the two divergences to avoid the formation of undesirable patterns in the process of learning clear pictures. Theoretical analysis proves that when an optimal discriminator is given, the difference between the forward and reverse KL divergences between real and generated data can be minimized. Experimental results demonstrate that compared to the existing methods, the proposed method can restore the details of an image more realistically and provides better performance in terms of the evaluation indexes of peak signal-to-noise ratio and structural similarity.
Activity Recognition Approach Using a Low-Resolution Infrared Sensor
ZHANG Yutong, ZHAI Xuping, WANG Jing
2022, 44(1): 47-53.
Abstract HTML (40) PDF(47)
Abstract:
The worldwide problem of population aging is becoming increasingly critical. To avoid accidents involving the elderly living alone, the study of the daily activities of the elderly using recognition and monitoring algorithms has become a research hotspot. This paper proposes an action recognition approach using a low-resolution infrared sensor. The proposed approach uses an infrared sensor to collect temperature distribution data in the detection area, and then processes the temperature distribution data, extracting multiple features in the four dimensions of time, temperature, deformation, and trajectory. Finally, the K-nearest neighbors algorithm is used to identify the five poses of "walking, " "bending, " "sitting, ""standing, " and "falling." Experimental results demonstrate that the average accuracy can reach 97% and that the accuracy for falling is 100%.
Systems & Designs
Design of a Three-Axis Turntable Servo Control Using Wirerope Transmission
WANG Kai, KONG Dejie, SONG Yueming, WANG Guohua, SHEN Honghai, WANG Hao
2022, 44(1): 54-60.
Abstract HTML (37) PDF(21)
Abstract:
Aiming at the problem of nonlinear errors and external disturbances caused by wirerope transmission in an airborne platform directly affecting the imaging quality of platform cameras, we propose a control strategy using fuzzy adaptive feedforward compensation. First, the wirerope transmission mechanism and a high-precision DC servo motor were modeled, a friction model was established, and a fuzzy adaptive proportional-integral-derivative controller was introduced for the speed loop control loop of the turntable. Based on the principles of feedforward compensation, a feedforward compensation algorithm is proposed. A compound control strategy combining feedforward compensation with a fuzzy adaptive controller is also designed. MATLAB simulation results and experiments demonstrate that this compound strategy can effectively eliminate the jitter caused by aircraft vibration and wirerope nonlinearity. The stability accuracy of the turntable is increased from 1 to 0.2 mrad, which can significantly improve imaging quality.
Topology Optimization Design and Analysis of an Integrated Aluminum Alloy Mirror
WANG Shang, ZHANG Xingxiang, SHA Wei, ZHU Junqing
2022, 44(1): 61-65.
Abstract HTML (73) PDF(39)
Abstract:
Topology optimization design was conducted for an integrated aluminum alloy mirror with a diameter of 300 mm. Under the self-weight load along the optical axis of the mirror, the global flexibility was considered as a constraint, and the minimum volume of the mirror was considered as an objective for iterative optimization to obtain a topology optimization model. According to the results, a solid model was established, and its parameters were optimized. Finally, an integrated mirror structure with a total mass of 2.08 kg, root mean square of 5.9 nm, and lightweight ratio of 70% was obtained. Through comparisons to a contrast structure combined with a parameter optimization process, the validity of the topological structure features was determined, and the support characteristics were analyzed. A support structure consisting of a central hexagon and semi-closed structure contributes significantly to the improvement of surface shape accuracy under the conditions of self-weight. There is an optimal supporting position for the central hexagon structure, where the ratio of height to diameter of the regular hexagon is 0.26.
Research on High-Speed Data Transmission Model of Large-Format High-Frame-Rate Readout Integrated Circuit
YE Lianhua, LIU Xu, LI Yunduo, HUANG Songlei, HUANG Zhangcheng
2022, 44(1): 66-72.
Abstract HTML (51) PDF(30)
Abstract:
In this paper, a high-speed data transmission model is presented for the digital signal output of a large-format high-frame-rate readout integrated circuit. We utilize the lumped parameter model to investigate the relationship between 3 dB bandwidth, response time, and device parameters. It is indicated that the size of the driver and the load of the transmission bus are the key parameters that determine the high-speed time-domain response characteristics. Furthermore, by using the distributed parameter model and Elmore delay model, the analytic expression for response time is deduced with more accurate values, and the optimal design of the output stage toward maximum bandwidth is obtained. Under the typical constraint condition of layout and power dissipation in a 64×64 array, simulation results show that the output 3 dB bandwidth of the transmission gate and composite logic gate can reach 293 MHz and 395 MHz, respectively.
Independent Design and Temperature Control Performance Experiment of the CdZnTe Crystal Growth Furnace
LUO Yanan, CHEN Yixin, GUO Guanzhu, LI Zhaocun, XU Cong
2022, 44(1): 73-78.
Abstract HTML (51) PDF(35)
Abstract:
In response to the demand for the growth of 4-inch diameter single-crystal CdZnTe materials and based on the results obtained from studying the growth of foreign CdZnTe crystal materials, a CdZnTe crystal growth furnace based on the mobile heating method was independently designed. The heating unit of the furnace body comprises four specifications with six temperature controlled sections, which are controlled by an industrial computer that controls the servo motor to drive the ball screw linear guide to achieve lifting. The inner cavity of the furnace was fitted with a heating tube comprising corundum ceramic tubes and high-temperature metal heat pipes using high-precision platinum-rhodium-platinum thermocouples, Eurofins, transformers, and thyristor control heating units. This furnace is based on a fuzzy + PID control algorithm with a strategy to adjust and control the temperature distribution of the heating furnace. This furnace was used to perform stability and control performance experiments during temperature heating. Experimental results showed that the heating temperature of the inner cavity of the furnace was continuously controlled for 200 h, temperature fluctuation at the same position was ±0.005℃, and heating temperature deviation was ≤ ±0.1℃. The lengths of the upper and lower constant-temperature zones of the furnace cavity were 400 and 240 mm, respectively. The length of the temperature gradient zone in the middle of the furnace cavity was approximately 136 mm. The length of the constant-temperature zone in the lower part of the furnace cavity was 240 mm. At a heating temperature of approximately 1098℃, the temperature gradient was 0.92℃⋅mm−1. Experimental results showed that this furnace meets the independent design and temperature control performance requirements for a CdZnTe crystal growth furnace.
Materials & Devices
Preparation of +4-Valent Vanadium Oxide Films via the Co-Sputtering of Mg and V2O5
FU Xuecheng, WU Liying, QUAN Xueling, QU Minni, WANG Ying
2022, 44(1): 79-84.
Abstract HTML (50) PDF(20)
Abstract:
Vanadium oxide films were deposited on a silicon substrate via co-sputtering a high-purity magnesium and vanadium pentoxide target under high vacuum at room temperature. Owing to the reducibility of the magnesium atom, the valency of vanadium was reduced from +5 to +4. When the atomic ratio of Mg to V was 1:2, X-ray diffraction (XRD) results showed that the main component in the film was MgV2O5. X-ray photoelectron spectroscopy (XPS) results showed that both V4+ and V2+ were present in the fabricated films. The crystallization of the MgV2O5 film was in adequate condition, as observed in the SEM profiles. The results of the temperature resistivity test showed that the film experienced a phase transition near room temperature. The hysteresis loop occurred at a temperature of approximately 0.3℃ with a temperature coefficient of resistance of −8.6%/K. The material constant of the negative temperature coefficient thermistor was approximately 6700. This discovery provides a novel method for the preparation of thermal film materials for application in uncooled focal plane detectors.
Surface Quality of InSb Wafer Using Different Polishing Methods
LI Dexiang, GONG Xiaoxia, ZHANG Lixia, YUAN Jun, YANG Wenyun
2022, 44(1): 85-88.
Abstract HTML (42) PDF(42)
Abstract:
In this study, InSb wafers were subjected to pure mechanical polishing and chemical mechanical polishing using hydrogen peroxide. The surface quality of the wafers polished using these methods was investigated by characterizing the surface scratch, surface roughness, and surface damage and verifying if an oxide film was formed on the wafer surface. The results showed that an oxide layer was formed on the wafer surface during mechanical polishing. This layer protects the material surface from damage. Moreover, chemical mechanical polishing using hydrogen peroxide as the polishing solution facilitated the formation of InSb wafers with good surface quality. The surface roughness was reduced to 0.606 nm, the flatness was about 6.916 nm, and the damage was significantly reduced.
Reliability Research for 640×512 Miniaturized IR Detector Dewar Assembly
XIONG Xiong, DUAN Yu, HU Mingdeng, LI Ruiping, DU Yu, MAO Jianhong
2022, 44(1): 89-95.
Abstract HTML (122) PDF(188)
Abstract:
Driven by the concept of SWaP3 (Size, Weight, and Power, Performance and Price), the development of the third-generation cooled IR detectors is proceeding in the direction of high performance, miniaturization, and light weight. As core military electronic devices, the reliability of IR detectors has become the focus of research. In this study, based on the 640×512/15 μm miniaturized dewar developed by Zhejiang Juexin Microelectronics Co., Ltd., a systematic reliability research is carried out. This research involves four dimensions, namely mechanics, thermodynamics, remainders, and vacuum. The performance of the 640×512/15 μm miniaturized dewar is evaluated through reliability tests. The results show that the miniaturized dewar has high reliability to satisfy most military needs.
Refrigeration
Research on Positive Semi-Definite Vibration Damping System of Single-Piston Linear Stirling Cryocooler
KONG Derui, XIA Ming, TANG Tianmin, BI Xiang
2022, 44(1): 96-102.
Abstract HTML (40) PDF(58)
Abstract:
At present, research on the vibration system of the vibration absorber and the single-piston linear Stirling cryocooler is almost entirely carried out on the positive definite model of a certain fixed installation method for the known cryocooler system. Owing to the lack of analysis of the system's natural frequency using this method, resonance may occur, and the damping effect of the vibration absorber may be reduced when the installation method is changed. Therefore, this study conducts a theoretical analysis of the positive semi-definite model of the cryocooler and the vibration system of the vibration absorber without any installation and performs vibration experiments before and after the cryocooler is fitted with the vibration absorber through the suspension test method. The theoretical natural frequency value of the system under the positive semi-definite model is 78.6 Hz. Through experiments, it can be found that, due to the actual processing and assembly of the vibration absorber, the natural frequency of the system changes with the vibration absorber's stiffness within the range of 78.1 to 80.8 Hz.

Copyright © 《Infrared Technology》Editorial Office京ICP备000000000号-1

Supported by: Beijing Renhe Information Technology Co. Ltd