首页 | 本学科首页   官方微博 | 高级检索  
     

基于单元复形的数字图像多级别层次表达
引用本文:朱国宾,刘小利,李清泉. 基于单元复形的数字图像多级别层次表达[J]. 地理与地理信息科学, 2008, 24(6)
作者姓名:朱国宾  刘小利  李清泉
作者单位:武汉大学国际软件学院,湖北,武汉,430079;武汉大学测绘遥感信息工程国家重点实验室,湖北,武汉,430079
基金项目:国家自然科学基金  
摘    要:基于复形理论定义了数字图像空间的拓扑元素及其性质,在此基础上提出一套完备的保持拓扑等价性的层次表达数字图像的数学模型体系框架,并验证了层次表达结构中的Jordan曲线定理.同时,基于单元复形扩展模型,对SPOT影像实施了渐进式离散分割,有效地利用影像蕴含的空间信息,获得了比最大似然分类法更优的分割结果.

关 键 词:数字图像  数字拓扑  抽象单元复形  数字Jordan曲线定理  图像表达

Multi-level Hierarchical Representation for Digital Images Based on Cell Complexes
ZHU Guo-bin,LIU Xiao-li,LI Qing-quan. Multi-level Hierarchical Representation for Digital Images Based on Cell Complexes[J]. Geography and Geo-Information Science, 2008, 24(6)
Authors:ZHU Guo-bin  LIU Xiao-li  LI Qing-quan
Affiliation:ZHU Guo-bin1,LIU Xiao-li2,LI Qing-quan2
Abstract:It's necessary to discuss the topology model of digital images for remote sensing and GIS.Based on complex theories,basic topological issues about digital image are given in this paper.A multi-level hierarchical image representation that preserves topological relation equivalency and a set of well-complete functional architecture that efficiently reflects this representation are presented,and the digital Jordan curve theorem is validated.In order to put these hierarchical ideas into computational practice,a novel progressive discrete segmentation algorithm is implemented that connected regions and their topology generate objects in different scales,furthermore a tree-object structure reflecting their spatial relationships is constructed.
Keywords:digital images  digital topology  abstract cell complexes  the digital Jordan curve theorem  image representation
本文献已被 CNKI 维普 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