基于图像特征的烟幕干扰效能评估方法  

A method for evaluating the effectiveness of smokescreen jamming based on image features

作　　者：刘书信[1,3] 丁佳麟 赵凤 陈春生 LIU Shu-xin;DING Jia-lin;ZHAO Feng;CHEN Chun-sheng(Southwest Institute of Technical Physics,Chengdu 610041,China;Chemical Defense Institute,Academy of Military Sciences,Beijing 102205,China;College of Optoelectronic Engineering,Chongqing University,Chongqing 400044,China)

机构地区：[1]西南技术物理研究所,四川成都610041 [2]军事科学院防化研究院,北京102205 [3]重庆大学光电工程学院,重庆400044

出　　处：《激光与红外》2025年第4期607-614,共8页Laser & Infrared

摘　　要：烟幕以其施放方式简单、效费比高等优点成为了现代战场的主要干扰手段之一,但目前针对烟幕干扰效能评估的方法较少,且缺少定量评估手段。本文提出了基于图像特征的烟幕干扰效能评估方法(HD-EEMSSJ),通过外场试验获取了真实试验数据并对该方法评估效果进行测试。该方法从导引头跟踪机制出发,将图像的方向梯度直方图特征、深度特征、余弦相似度以及亮度特征进行加权融合,得出定量化的评估结果HD-EEMSSJ指数,该指数能够更加准确且敏锐地体现出烟幕的动态干扰情况,为后期干扰效果的分级评定提供参考依据。经多组试验数据验证结果表明,HD-EEMSSJ指数对比传统的图像质量评估方法PSNR、RFSIM、SSIM以及本文提出的EEMSSJ方法具有更好的评估效果,准确性分别提高了533.15%、170.2%、26.4%和3.25%。Smoke screen has become one of the main jamming methods in modern battlefields due to their advantages of a simple deployment method and a high cost-effectiveness ratio.However,there are fewer methods for assessing the jamming effectiveness of smoke screen and lack of quantitative assessment means.In this paper,a method for evaluating the effectiveness of smoke screen jamming based on image features(HD-EEMSSJ,HOG&Depth-Effectiveness evaluation of smoke screen jamming)is proposed,and real experimental data are obtained and the evaluation effect of the method is tested through the field test.Starting from the guide tracking mechanism,the direction gradient histogram features,depth features,cosine similarity and brightness features of the image are weighted and fused to obtain a quantitative evaluation result HD-EEMSSJ index,which can more accurately and keenly reflect the dynamic interference of the smoke screen,and provide a reference basis for the grading of the interference effect in the later stage.The results of multiple sets of experimental data verification show that the HD-EEMSSJ index provides better assessment than the traditional image quality assessment methods PSNR,RFSIM,SSIM and the EEMSSJ method proposed by the authors,with accuracy improvements of 533.15%,170.2%,26.4%and 3.25%,respectively.

关 键 词：图像特征 烟幕干扰 效能评估 方向梯度直方图 

分 类 号：TN976[电子电信—信号与信息处理] TN219[电子电信—信息与通信工程]