文章基本信息

标题：A Modified Segmentation Approach for Synthetic Aperture Radar Images on Level Set
本地全文：下载
作者：Peng, Qiangqiang ; Zhao, Long
期刊名称：Journal of Software
印刷版ISSN：1796-217X
出版年度：2013
卷号：8
期号：5
页码：1168-1173
DOI：10.4304/jsw.8.5.1168-1173
语种：English
出版社：Academy Publisher
摘要：In this paper, we propose a method towards unsupervised segmentation of synthetic aperture radar (SAR) image for segmentation of homogeneous regions. The SAR amplitude image is modeled by Rayleigh distribution. Moreover, in order to avoid geometric topology variations, the curve evolution is implemented via level sets. In addition, a new energy function is proposed, which is more objective and robust for the segmentation purpose. During the curve evolution, a convex factor is considered to guarantee the regularity of the region boundary. Finally, a novel termination criterion of the evolution of energy function is designed. And then, the effectiveness of our method is demonstrated on real SAR images.
关键词：Synthetic Aperture Radar Imagery;Segmentation;Active Contours;Level Set