利用DT-GrowCut的MSTAR SAR图像自动分割技术  

SAR image automatic segmentation based on DT-GrowCut

作　　者：崔艳鹏[1] 胡建伟[1] 杨绍全[1] 朱燕[1] 

机构地区：[1]西安电子科技大学电子对抗研究所,陕西西安710071

出　　处：《西安电子科技大学学报》2011年第3期150-154,共5页Journal of Xidian University

基　　金：西安电子科技大学基本科研业务费资助项目(72003052);国家部委预研基金资助项目(J14406020163);国家部委预研基金资助项目(9140A07020806DZ01)

摘　　要：结合Delaunay三角剖分理论,提出了一种DT-Growcut全自动SAR雷达图像分割技术.首先将MSTAR数据图像进行Delaunay三角剖分.由于背景噪声是随机相干斑噪声,所以选择两个最大的连通域,引导设置GrowCut种子函数,依据自动细胞机竞争机制,对SAR图像进行分割处理.该方法不需要预先设置类别,能够消除相干斑噪声,能有效地提取SAR图像边沿,大大降低图像的边缘模糊.通过对MSTAR数据库进行仿真实验,并对分割结果进行分析,证明了该方法的有效性和可行性.A novel automated robust segmentation algorithm based on Delaunay triangulation and GrowCut is proposed. The core idea is to apply the theory of Delaunay triangulation to the MSTAR image just for the seeds initialization of GrowCut. The DT-GrowCut based algorithm can give rough segmentation between the target, shadow and background. Several sets of SAR images from MSTAR datasets are segmented by the proposed method. Experimental results demonstrate that the proposed method outperforms the traditional FCM based method in despeckling, extracting the edge and reducing the edge fuzzy degree. The results of experiments with different SAR images demonstrate the efficiency of the presented algorithm.

关 键 词：SAR图像分割 DELAUNAY三角剖分 GrowCut MSTAR 

分 类 号：TN911.7[电子电信—通信与信息系统]