首页 | 本学科首页   官方微博 | 高级检索  
     检索      

散射特性引导的极化SAR图像多尺度分割
引用本文:巫兆聪,欧阳群东.散射特性引导的极化SAR图像多尺度分割[J].测绘科学,2012,37(3):35-37.
作者姓名:巫兆聪  欧阳群东
作者单位:武汉大学遥感信息工程学院,武汉,430079
基金项目:国家863计划资助项目,国家自然科学基金资助项目
摘    要:为满足不同地物分割的尺度需求,本文提出一种散射特性引导的极化SAR图像多尺度分割算法。该算法以分水岭初始分割为基础,通过迭代合并相邻区域获取极化SAR图像在不同尺度下的分割结果;以极化散射功率为特征,将地物按尺度相似性划分成表面散射、二次散射和体散射3种散射类型;以分割质量评价为指导,从多尺度分割结果中为每种散射类型地物选取适宜的分割尺度并进行叠置合成,形成最终分割结果。ESAR系统获取的L波段极化SAR数据验证了该算法的可行性。

关 键 词:极化SAR  散射特性  多尺度分割

Multi-scale segmentation of polarimetric SAR imagery guided by scattering properties
WU Zhao-cong , OUYANG Qun-dong.Multi-scale segmentation of polarimetric SAR imagery guided by scattering properties[J].Science of Surveying and Mapping,2012,37(3):35-37.
Authors:WU Zhao-cong  OUYANG Qun-dong
Institution:WU Zhao-cong,OUYANG Qun-dong(School of Remote Sensing and Information Engineering,Wuhan University,Wuhan 430079,China)
Abstract:Considering the different segmentation scales required for various ground objects,a novel multi-scale segmentation method guided by scattering properties was proposed in this paper.To implement this method,it firstly obtained the initial segmentation results with different scales by iteratively merging regions,which are derived from watershed transformation.Then,ground objects were divided into surface,double and volume scattering categories in terms of their segmentation scales,using the feature of polarimetric scattering powers.Finally,the desired segmentation result was achieved by selecting and synthesizing suitable scales for each scattering category,according to segmentation quality evaluation.The effectiveness of this method was demonstrated using L-band polarimetric data acquired by ESAR system.
Keywords:polarimetric synthetic aperture radar  scattering property  multi-scale segmentation
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号