一种评价红外变形伪装效果的新方法  

A Quantitative Evaluating Method for Infrared Distortion Camouflage Effect

作　　者：张健[1] 赵保军[1] 毛二可[1] 

机构地区：[1]北京理工大学电子工程系

出　　处：《激光与红外》2006年第1期68-71,共4页Laser & Infrared

摘　　要：文中提出了一种新的评价红外伪装效果的方法。该方法主要应用Moravec算子提取军事目标变形伪装前、后红外图像的特征点,用改进的Hausdorff距离计算特征点集之间的Hausdorff距离。该距离越大,说明军事目标经过伪装后的变形效果越好。另一个方面,通过把变形伪装后的目标及其背景的红外图像转换为模糊集,用基于模糊集M inkowski距离的方法,计算出目标与其相邻背景红外辐射特性的相似度。目标与背景红外图像的M inkowski距离越小,则说明目标变形伪装后的红外伪装效果越好。通过这个两个方面的计算,可以定量评价出目标变形伪装的红外伪装效果。In this paper,we propose a new method to evaluate infrared distortion camouflage effect. We adopt Moravec operator to extract the feature points of the images of target taken before and after it is camouflaged, and then we use the improved Hausdorff distance to measure the Hausdorff distance between the feature points sets. The greater the distance is ,the better the camouflage effect is. The other necessary aspect evaluating the camouflage effect is to measure the similarity of the infrared radiation character between the camouflaged target and its adjacent background. The similarity measure is based on the fuzzy Minkowski distance, so we have to transform the images into fuzzy sets at first. The observation is that the smaller the distance between the infrared image of the camouflaged target and its adjacent background is ,the greater the similarity is between them and the better the infrared camouflage effect is, In the two ways,we can evaluate the infrared distortion camouflage effect of the camouflaged target quickly.

关 键 词：伪装 变形 效果 评价 距离 

分 类 号：TN976[电子电信—信号与信息处理]