最大均匀区域滤波器在SAR图象处理中的应用

Applications of maximum homogeneous region filter to SAR image processing

介绍了最大均匀区域滤波器的基本原理 ,为获得预期的滤波效果 ,在实现过程中 ,采用模板匹配的方法进行阶跃边缘的检测 ,使窗口移动的方向更易确定 ;用聚类分析与统计特征参数相结合的方法 ,建立不同灰度区域的判别门限 ,使门限值更加准确 ;通过重新估计噪声的相对标准差 ,对局部滤波算法进行了改进。实验结果表明 ,这些方法和措施是有效的。

The principle of maximum homogeneous region filter is introduced. In order to obtain the expected filtering effect,the template matching method is used to realize the step edge detection, which can decide the direction of the window movement more easilly The more accurate thresholds to distinguish can be set up by combining cluster analysis with statistics paramenters, the thresholds are more accurate. By re-estimating the standard deviation of the speckle, the performance of the local filtering algorithm is improved. The final experiment results indicate that the methods are effective.