基于卡尔曼滤波的SAR图像边缘检测方法  

Edge Detection for SAR Images Based on Kalman Filter

作　　者：李占利[1] 刘宇琦 孙瑜[1] 李洪安[1] 张蕴[1] LI Zhan-li;LIU Yu-qi;SUN Yu;LI Hong-an;ZHANG Yun(College of Computer Application Technology,Xi’an University of Science and Technology,Xi’an Shaanxi 710054,China)

机构地区：[1]西安科技大学计算机科学与技术学院

出　　处：《图学学报》2019年第5期823-828,共6页Journal of Graphics

基　　金：陕西省教育厅科研计划项目(17JK0513);陕西省自然科学基础研究计划项目(2019JM-162);西安科技大学博士启动金项目(2019QDJ007)

摘　　要：针对传统Canny边缘检测算法对合成孔径雷达(SAR)图像的相干斑噪声抑制程度太高,导致大量边缘的真实信息丢失问题,提出一种新型Canny算子边缘检测算法。首先建立合适的非对称半平面区域(NSHP)图像模型,将空间模型转换成卡尔曼滤波可适用的系统状态方程;然后用"预测+反馈"的方式对图像去噪;最后通过双阈值算法提取图像的边缘。仿真实验表明,该方法可以有效地抑制SAR图像中的相干斑噪声,同时能较好地保留图像的边缘信息,相对于传统的Canny算法有较好的检测效果。Traditional Canny edge detection algorithm suppresses the speckle noise of synthetic aperture radar(SAR) images too much, causing much loss of real edge information loss. To tackle this problem, this paper proposed a new Canny operator edge detection algorithm. Firstly, the method established a suitable non-symmetric half plane(NSHP) image model, then converted the spatial model into a system state equation applicable to Kalman filter;after that, then we adopted the method of prediction and feedback to denoise the image. Finally, the edge of the image was extracted by dual threshold algorithm. Experimental results show that the proposed method can effectively suppress the speckle noise of the SAR image and preserve the edge information well, and provide better detection effects than traditional Canny algorithm.

关 键 词：CANNY算子 边缘检测 NSHP 卡尔曼滤波 SAR图像 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]