Research On Target Infrared Polarization Characteristics And Image Enhancement Technology In The Background Of Figures

Infrared detection technology is widely used in the military combat field because of its advantages of all-weather operation,long acting distance and strong anti-interference capability.However,for low or non-heat emitting targets,such as UAVs and stationary vehicles,the temperature difference between the target and the background is small and the infrared characteristics are not obvious leading to the difficulty of accurate target identification using traditional infrared detection technology.Thus,how to effectively detect faint targets from complex backgrounds is a hot issue that needs to be addressed urgently.Infrared polarization detection technology,based on the difference of polarization characteristics of target and background,can provide a “new dimension” of information different from the intensity and spectrum,thus providing an effective solution to the detection and identification of targets in complex backgrounds.This paper systematically investigates the problem of target infrared polarization characteristics and image enhancement in the ground background,focusing on the problems of discrepancy between the polarization characteristics modeling and the actual microplane elements,and the lack of in-depth integration of the polarization characteristics of target and background,etc.The study is carried out in three aspects: infrared polarization characteristics modeling from the physical mechanism,simulation of the target infrared polarization radiation scene,and integration of infrared and polarization information with infrared polarization imaging technology.The specific contents are summarized as follows:1.To address the problem that the infrared polarization characteristics of the target are not consistent with the actual surface elements,the two-component pBRDF model of the target surface is proposed based on specular reflection and diffuse reflection,and the infrared polarization degree calculation model of the target is derived.The influence of several factors on the polarization characteristics of the target coating is analyzed by numerical results.The results show that the greater the roughness of the target surface,the more obvious the shading effect caused by the roughness,the smaller the infrared polarization of the corresponding target;the ambient radiation of the target coating is negatively correlated with its polarization.2.For the problem that the simulation of infrared characteristics of the ground target takes less consideration of polarization factors,the infrared and polarization characteristics of different ground targets are simulated based on infrared radiation theory and polarization degree model,respectively.The geometric models of the tank and the car are established by using3 Dmax,and Matlab is used to program the fading judgment,coordinate transformation,and visualization of infrared radiation and polarization.The differences between the IR radiation and polarization simulation images of the tank and the car in terms of target IR information highlighting are discussed separately.The results show that compared with the infrared radiation intensity image,the contrast and details of the infrared polarization image are better,so it is necessary to further analyze the polarization feature image of the target.3.To address the problem of insufficient depth of integration between the feature background and the target IR polarization features,a passive time-sharing IR polarization imaging system is used to acquire,pre-process and enhance the IR and polarization images of targets in different feature backgrounds based on the principle of IR polarization imaging.The polarization images of the target and background are extracted separately based on the Stokes vector method,and the difference in polarization between them is analyzed using histogram normalization.Two fusion algorithms are used to fuse the infrared image and polarization image,and the results are evaluated by objective evaluation metrics.The results show that the subjective and objective evaluation metrics are significantly improved compared with those before fusion.Therefore,the fusion enhancement of the target image is achieved.