基于纹理的SAR图像居民地信息提取  

The Information Extraction of Residential Area from SAR Image Based on Texture

作　　者：李海权[1] 胡召玲[1] 孙陟文[1] 张国成[1] 

机构地区：[1]徐州师范大学城市与环境学院,徐州221116

出　　处：《遥感信息》2008年第2期56-59,79,共5页Remote Sensing Information

基　　金：江苏省高校自然科学研究计划项目(05KJB420133)

摘　　要：研究了利用纹理特征来实现居民地专题信息的提取。首先用增强FROST滤波算法对SAR图像进行去噪处理。然后基于灰度共生矩阵实现纹理特征的提取,结合实际现状筛选出较好的纹理特征图像。最后将各图像进行波段组合实现非监督分类。实验结果表明:基于纹理特征来提取SAR图像中的居民地信息是可行的,效果较理想。The thematic information of residential area extracted by using texture features is researched in this paper. First the speckle from SAR image is eliminated by the enhanced FROST filter algorithm . Then texture features are extracted based on gray-level co-occurrence matrix. The better images of texture features are selected according to actual situation. At last, the selected images are combined into a image with many bands to perform non-supervised classification. Experimental results illustrate that the information extraction of residential area from SAR image based on texture features is feasible and effect is more ideal.

关 键 词：纹理 灰度共生矩阵 SAR图像 居民地 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]