西安电子科技大学学报（自然科学版）2011，Vol.38Issue(3)：150-154,5.DOI:10.3969/j.issn.1001-2400.2011.03.024
利用DT-GrowCut的MSTAR SAR图像自动分割技术
SAR image automatic segmentation based on DT-GrowCut
摘要
Abstract
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/MSTARKey words
synthetic aperture radar image segmentation/ Delaunay tringulation/ GrowCut/ moving and stationary targets acquisition and recognition分类
信息技术与安全科学引用本文复制引用
崔艳鹏,胡建伟,杨绍全,朱燕..利用DT-GrowCut的MSTAR SAR图像自动分割技术[J].西安电子科技大学学报（自然科学版）,2011,38(3):150-154,5.基金项目
西安电子科技大学基本科研业务费资助项目(72003052) （72003052）
国家部委预研基金资助项目(J14406020163) （J14406020163）
国家部委预研基金资助项目(9140A07020806DZ01) （9140A07020806DZ01）
