极化SAR图像解译中的不相似性度量  

Dissimilarity Measures in PolSAR Image Interpretation

作　　者：潘婷 杨祥立 宋辉[1] 杨文[1] PAN Ting;YANG Xiangli;SONG Hui;YANG Wen(School of Electronic Information, Wuhan University, Wuhan 430072, China)

机构地区：[1]武汉大学电子信息学院

出　　处：《无线电工程》2019年第8期649-656,共8页Radio Engineering

基　　金：国家自然科学基金资助项目(61771351)

摘　　要：不相似性度量常用来描述极化合成孔径雷达(SAR)样本数据之间的差异性,已被广泛应用于滤波、目标检测、分割分类和变化检测等多种极化SAR图像解译技术中。基于极化SAR数据的矩阵表达形式,从特征、统计分布、假设检验、信息论散度和信息几何等多角度出发,总结回顾了极化SAR图像解译问题中常见的不相似性度量,简要分析了其优势、存在的问题以及适用范围,并对极化SAR不相似性度量的研究趋势进行了展望。Dissimilarity measures are often used to describe the difference between the polarimetric SAR(PolSAR) samples,which have been widely used in various PolSAR image interpretation techniques,such as filtering,object detection,segmentation,classification and change detection.Based on matrix expression form for PolSAR data,the dissimilarity measures in PolSAR image interpretation is summarized and reviewed from such aspects as feature,statistical distribution,hypothesis test,information theory divergence,information geometry and so on.The advantages,problems and applications of these dissimilarity measures are briefly analyzed.Finally,The future study of PolSAR dissimilarity measures is pointed out.

关 键 词：极化 合成孔径雷达 距离测度 不相似性 协方差矩阵 

分 类 号：TP957.52[自动化与计算机技术]