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

一种新型的图象形状松弛匹配方法
引用本文:陈梁,翟羽健.一种新型的图象形状松弛匹配方法[J].东南大学学报(自然科学版),1988(4).
作者姓名:陈梁  翟羽健
作者单位:南京工学院自动控制系,南京工学院自动控制系
摘    要:松弛匹配方法是一种重要的图象匹配方法。本文通过图象中特征间的相互关系,利用松弛过程不断更新标识符的原始值,并且引入简单方便的寻优技术,以减少混乱度及提高相容性为原则提出了一种优化准则,从而加速及改善了松弛过程。本文提出的松弛匹配方法还可应用于识别重叠景物,结果优于一般的松弛匹配方法。

关 键 词:松弛匹配  单纯形寻优  标识符  混乱度

A New Kind of Image Shape Relaxation Matching Method
Chen Liang Zhai Yujian.A New Kind of Image Shape Relaxation Matching Method[J].Journal of Southeast University(Natural Science Edition),1988(4).
Authors:Chen Liang Zhai Yujian
Institution:Department of Automatic Control Engineering
Abstract:Relaxation method is an important image shape matching method. In this paper, relaxation process is used repeatly to update the values of the probability of labels which are depended on the relations between image features. A simple and convenient optimization method is introduced and a optimization standard is proposed in order to reduce ambiguity and to raise consistency. Experimental results show that the relaxation process has been improved. The method can also be used in over lapped object recongnition, and the result is better than that of the ordinary one.
Keywords:relaxation matching  simplex method  label  ambiguity  
本文献已被 CNKI 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