一种改进的SAR图像边缘检测方法

一种改进的SAR图像边缘检测方法 An Improved Edge Detection Algorithm of SAR Images 吕毅 L Yi 1 2 first-author 孙洪 SUN Hong 1 2 曹永锋 CAO Yong\|feng 1 2 王青 WANG Qing 1 2 徐戈 XU Ge 1 2 武汉大学电子信息学院 武汉大学电子信息学院 School of Electronic Information, Wuhan University School of Electronic Information, Wuhan University 武汉大学电子信息学院,湖北,武汉,430072 武汉大学电子信息学院,湖北,武汉,430072 针对Ratio边缘提取算法对弱边缘检测效果差 ,边缘定位精度低的缺点 ,提出了一种改进算法用于检测微弱边缘 ,并给出具有CFAR特性的阈值选取办法 实际应用中 ,将两种方法结合为一种综合处理方法 ,对真实SAR图像的处理结果表明 ,该综合处理方法具有较高检测正确率及较好的边缘定位精度 Ratio edge detector for SAR images can provide high overall edge\|detection probability. But it has defects on the accuracy of edge\|position determination. This paper works deeply on how to determinate the threshold of Ratio edge detector. Then we present a new detector aiming at detecting “weak” edges with threshold of it. A new combined algorithm including both of the two detectors is presented. Its processing results of actual SAR images show wonderful edge\|detection performance and high accuracy of edge\|position determination. 合成孔径雷达 Speckle噪声 边缘检测 图像处理 synthetic aperture radar Speckle noise edge detection image processing TN957.52 国家自然科学基金资助项目 （60 0 72 0 4 1 ） 2003-01-01 2021-04-01 1