论文部分内容阅读
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。
Therefore, the removal of speckle noise has aroused general concern of microwave remote sensing scientists. There are many scholars successively Proposed a variety of different processing methods, of which the most commonly used mean smoothing, median smoothing, Frost adaptive filtering, Lee adaptive filtering and the recent by P. V. NarasimhaRao et al proposed improved K-means adaptive filtering and other methods. This paper analyzes the mathematical basis of these methods, and selects a representative experimental area to carry out comparative tests. The results show that the modified K-means adaptive filtering method is the best.