一种基于保留边缘的SAR图像相干斑滤除方法

研究一种能保持边缘的星载ＳＡＲ图像相干斑噪声滤除方法,常规滤波算法在滤除噪声的同时也带来了图像边缘模糊问题,利用多种检测图像边缘算法的有机组合,首先提取出图像边缘,然后将边缘从图像中剔除,对剩下的无边缘噪声图像用改进Ｌｅｅ滤波法进行噪声抑制,再把边缘图像加入,最后得到了能保持边缘的ＳＡＲ图像相干斑滤除图像。用实际数据进行检验,并和常规方法得到的结果比较,验证了本算法的有效性。

A Speckle Reduction Algorithm by Edge-Preserving for SAR Images

A speckle reduction by edge preserving method to solve spaceborne SAR image's edge ambiguity after normal speckle reduction was developed. Based on joint several edge detection methods, the edge was detected from SAR image by using Haralick algorithm and general fuzzy operator(GFO) algorithm, and then the detected edge was subtracted from SAR image, then Lee filter was used to filter the speckle image without edge, at last the edge was added to the filtered image. The new method was demonstrated by spaceborne SAR image. It is a good way for speckle reduction for SAR images.