Robustness study of gray distribution characteristics in true and false target recognition

Abstract

Setting false targets is one of the important means of battlefield camouflage. The survivability of protected targets largely depends on the effect of photoelectric deception and jamming of false targets. The comprehensive use of robust image features plays a key role in correctly evaluating the photoelectric deception jamming efficiency of false targets and quickly identifying true and false targets. The existing efficiency evaluation models lack systematic research on the robustness of various image features in different environments. There are common problems of single target background and incomplete consideration of environmental factors, which lead to the instability of the extracted target image features, and then affect the evaluation results.Given the above problems, two kinds of environmental conditions are set: the change of observation distance and the change of atmospheric attenuation intensity. The experimental environment is simulated by an equivalent simulation method. After filtering the target background interference and extracting the target subject image and features, the similarity measurement method is used to calculate the similarity between true and false target subject images. By comparing the changes of gray distribution, contour, and texture feature similarity value, the robustness ranking and change reasons of features under different environmental conditions are summarized. At the same time, experiments are designed to verify the robustness of gray distribution features. The results show that the gray distribution feature has strong robustness, and the combination of gray distribution feature on the battlefield can effectively help officers and soldiers identify true and false targets in a complex environment.