A Novel SAR Image Segmentation Method Using Run-Length Grouping  

作　　者：HU Lei JI Yan HOU Pengyang LI Bo 

机构地区：[1]Beijing Key Laboratory of Digital Media, School of Computer Science and Engineering, Beihang University, Beijing 100191, China [2]Beijing Remote Sensing Institute, Beijing 100085, China [3]School of Computer Information Engineering, Jiangxi Normal University, Nanchang 330022, China

出　　处：《Chinese Journal of Electronics》2013年第2期396-400,共5页电子学报（英文版）

基　　金：This China （No.61262036）, the National Basic Research Program of for Distinguished Young Scholars （No.61125206）. work is partially supported by the National Natural Science Foundation of China （973 Program） （No.2010CB327900） and the National Science Fund

摘　　要：Synthetic aperture radar （SAR） image seg- mentation is the basis of SAR image analysis and under- standing. A segmentation method based on the neigh- bor and near spatial relationship is proposed to reduce the speckle noises influence and improve the segmentation region regularity. The SAR image is coarsely segmented by the scatter intensity of neighbor pixels. For the un- regularity of region in coarse segmentation, a region reg- ularity method based on Run-length grouping （RLG） is proposed. The region regularity method adjusts the short run-length segmentation type with the spatial relationship of near pixels. The segmentation region can be more reg- ularity by connecting the broken region~ smoothing the tortuous edge and removing patch.

关 键 词：Synthetic aperture radar （SAR） image Image segmentation  Run-length grouping. 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN958[自动化与计算机技术—计算机科学与技术]