星载SAR图象斑点噪声消除方法效果的比较研究

Evaluation of Smoothing Filters Suppressing Speckle Noise on SAR Images

星载（ＳＡＲ）图象一般都存在相干斑点噪声，严重干扰地物信息的提取与ＳＡＲ图象的应用效果，因此斑点噪声的消除问题引起微波遥感科技工作者的普遍关注，先后有多位学者提出各种不同的处理方法，其中最为常用的有均值平滑滤波、中值平滑滤波、Ｆｒｏｓｔ自适应滤波、Ｌｅｅ自适应滤波以及近期由Ｐ．Ｖ．ＮａｒａｓｉｍｈａＲａｏ等人提出的改良Ｋ－均值自适应滤波等方法。本文分析了这几种方法的数学基础，并选择了有代表性的实验区开展对比试验，结果表明：改良Ｋ－均值自适应滤波法效果最佳。

Synthehc Aperture Radar (SAR) images are usually corrupted by speckle noise. In order to suppress the speclde noise and improve interpretation of the radar images, smoothing filters have been proposed in the literature. Among them, the most popular are used. Several types of filters are: the average filter, the median filter, the Frost adaptive filter, the tee adaptive filter. Recently, Rao et al. (1995) proposed a new method called the modified K-value adaptive filter. This paper is an intercomparison study of these filters using ERS-1 SAR data of the Washita area in Okalahoma, United States acquired on June 10,1992. Summaries of these filters are given. Results of applying these filters to the data are evaluated with two quantitative measures. Based on the intercomparison, the most appropriate filter of SAR image speckle noise in this study was the modified K-value adaptive filter.