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

一种准自动多传感器图像配准方法
引用本文:彭晓明,丁明跃,周成平,马茜.一种准自动多传感器图像配准方法[J].华中科技大学学报(自然科学版),2003,31(4):85-87.
作者姓名:彭晓明  丁明跃  周成平  马茜
作者单位:1. 华中科技大学图像识别与人工智能研究所图像信息处理与智能控制教育部重点实验室
2. 中国科学院光电技术研究所
基金项目:国家自然科学基金资助项目 (6 0 1 35 0 2 0FF0 30 4 0 5 ) .
摘    要:针对现有多传感器自动配准方法存在的应用局限性,提出了一种人机交互式粗—精结合的准自动配准方法.先选取少量控制点进行粗匹配,在此基础上利用整体图像信息进行精匹配.该方法将基于特征和基于区域的配准方法进行了有效结合,从而保证了配准的精度和效率,取得了令人满意的效果.此方法可以作为现有的自动配准方法的补充.

关 键 词:图像融合  多传感器  图像配准  准自动配准方法  图像匹配  图像处理
文章编号:1671-4512(2003)04-0085-03
修稿时间:2002年11月5日

A quasi-automatic method for multi-sensor image registration
Peng Xiaoming Ding Mingyue Zhou Chengping Ma Qian Peng Xiaoming Doctoral Candidate.A quasi-automatic method for multi-sensor image registration[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2003,31(4):85-87.
Authors:Peng Xiaoming Ding Mingyue Zhou Chengping Ma Qian Peng Xiaoming Doctoral Candidate
Institution:Peng Xiaoming Ding Mingyue Zhou Chengping Ma Qian Peng Xiaoming Doctoral Candidate, Institute of Pattern Recognition & AI,Huazhong Univ. of Sci. & Tech.,Wuhan 430074,China.
Abstract:A quasi automatic coarse to fine registration method was presented. The proposed method consisted of two steps: selecting several control point sets in a man machine manner, and using whole information of images for fine registration. It combined advantages of both feature based and area based methods and could guarantee both efficiency and accuracy of image registration. Experimental results showed its effectiveness. It could be used as a complement to the existing automatic methods for multi sensor image registration.
Keywords:image fusion  multi  sensor image registration  quasi  automatic registration method
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