一种基于复Wishart-Chernoff距离的极化SAR图像边缘检测算法  

An ROA Method of Edge Detection for SAR Imagery Based on Complex Wishart Distribution

作　　者：邵鹏 陈英 吴元 SHAO Peng;CHEN Ying;WU Yuan(Southwest China Institute of Electronic Technology,Chengdu 610036,China;Chengdu Technological University,Chengdu 611730,China)

机构地区：[1]中国西南电子技术研究所,成都610036 [2]成都工业学院,成都611730

出　　处：《电子信息对抗技术》2020年第1期16-20,45,共6页Electronic Information Warfare Technology

摘　　要：针对单极化SAR图像可用目标散射信息维度有限的问题,提出一种基于复Wishart-Chernoff距离的极化SAR图像边缘检测算法,充分利用极化协方差矩阵解决了单极化SAR图像边缘检测中边缘模型限制及边缘定位不准问题。该算法首先对极化协方差矩阵进行相干斑抑制,利用Wishart-Chernoff距离定量分析两个协方差矩阵的相似性,并搜索4个方向的局部极值,从而完成极化SAR图像的边缘检测。利用该方法处理了AIRSAR-Flevoland全极化数据,处理结果表明了该算法的有效性。A method is proposed to identificate active decoy target transmitted by repeater jammer. The difference of polarization characteristic for active decoy and real target are exploited. Active decoy and real target are classified, and active decoy is identificated. Polarization ratio is determined by the high probability decoy target. Real data is processed and results indicate that decoy target is discriminated.

关 键 词：边缘检测 SAR图像 协方差矩阵 复Wishart分布 Chernoff距离 

分 类 号：TN971[电子电信—信号与信息处理]