通过分析SAR影像的Speckle局部统计特性和现有的空间滤波算法 ,提出了一种新的Speckle滤波方法 ,该方法以局部相对标准差为主要依据 ,采用特殊的滤波窗口邻域划分方法 ,自适应地调整窗口尺寸和窗口内的滤波区域。该方法被用于多幅ERS 1/ 2SAR图像去Speckle处理 ,并与以往的典型算法进行了比较实验 ,取得了令人满意的结果。 By analyzing the local statistical characteristics of SAR images and the existing spatial filtering algorithms, a new Speckle filtering method is proposed. This method takes the local relative standard deviation as the main basis, uses the special filtering window neighborhood partitioning method, Adjust the window size and filter area within the window. The proposed method is applied to the de-Speckle processing of multiple ERS 1 / 2SAR images. Compared with the classical algorithms, the proposed method has achieved satisfactory results.