SAR图像滤波方法比较与分析  

The Comparison and Analysis of SAR Image Filtering Methods

作　　者：王宇航[1] 范文义[1] 张金虎[1] 

机构地区：[1]东北林业大学林学院,哈尔滨150040

出　　处：《森林工程》2015年第3期81-84,共4页Forest Engineering

基　　金：国家高技术研究发展计划(2012AA102001)

摘　　要：相干斑噪声是SAR系统的固有原理缺陷,其阻碍了SAR图像的后续分类应用。针对传统滤波方法在噪声去除及边缘保持方面的不足,提出适用于同质区域和边缘区域的自适应滤波方法对SAR图像进行滤波处理。首先描述SAR图像斑点噪声的产生机理及统计特性,其次根据图像后续分类的应用目的,对常用滤波器进行分析并将福建将乐林场RADARSAT-2图像数据分别进行LEE与增强LEE滤波、FROST与增强FROST滤波、GAMMA滤波、KUAN滤波、LOCAL SIGMA滤波处理,以有效视数、图像边缘保持指数等作为评价指标。最后将实验结果依据评价指标进行对比分析。结果表明,增强型LEE自适应滤波综合效果最好,能在较好去除斑点噪声的同时又保持图像的边缘信息。通过系统比较分析不同的滤波方法,从理论上为SAR图像后续森林类型分类应用前滤波方法的选择提供了理论依据。The speckle noise is an inherent defect in SAR system, which hindered the subsequent application of SAR images. To deal with the shortage of speckle removal and edge preservation in the SAR despeckle operations, an adaptive despeckle filter that can preserve edges based on noise identification and edge detection oriented to SAR images was proposed. In this paper, the mecha- nism of SAR image speckle noise and statistical properties were first described and then according to the purpose of image classifica- tion, the common speckle filters, including LEE and enhanced LEE filtering, FROST and enhanced FROST filtering, GAMMA fihe- ring, KUAN filtering, and LOCAL SIGMA filtering were applied to Jiangle State Forest Farm in Fujian RADARSAT-2 images. Final- ly, the results were analyzed based on the evaluation index. The results showed that the enhanced LEE adaptive filter was the best, which can maintain better image edge information and at the same time remove the speckle noise. The comparison and analysis of the system＇s different methods of filtering has provided a theoretical basis of selecting filtering method for the SAR image classification of forest types.

关 键 词：SAR 相干斑噪声 滤波处理 RADARSAT-2 

分 类 号：S771[农业科学—森林工程]