一种遥感图像的分割方法  

A segmentation method for remote sensing images

作　　者：宋晓东[1] 高勇[1] 阙大顺[1] 

机构地区：[1]武汉理工大学信息工程学院,湖北武汉430063

出　　处：《西部探矿工程》2003年第3期182-184,共3页West-China Exploration Engineering

摘　　要：图像分割是遥感图像处理中很重要的一步。因SAR图像通常带有较强的噪声 ,用传统的边缘检测方法效果不理想。利用数学形态学开闭运算和混合滤波 ,可据目标的形状选用算法中的探针 ,取得较好的滤波去噪和目标分割的效果。Image segmentation is an important part of the processing RS(remote sensor) images. Traditional methods of edge detection can't get a satisfied result on RS images, because they always have a lot of noise. The mathematical morphological mixture filters composed of Opening and Closing algorithm is proposed in this paper. The important point is that structure elements can be selected flexibly. Result show that mathematical morphology is effective in this field.

关 键 词：遥感图像 分割方法 图像分割 数学形态学 SAR 噪声 滤波 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]