Computer Engineering and Applications ›› 2006, Vol. 42 ›› Issue (19): 20-22.

• 博士论坛 • Previous Articles     Next Articles

Theory and Experiment of Registration Similarity Measure based on Fuzzy Mutual information

HaiNan Wang,   

  1. 西北工业大学
  • Received:2006-04-06 Revised:1900-01-01 Online:2006-07-01 Published:2006-07-01
  • Contact: HaiNan Wang



  1. 西北工业大学
  • 通讯作者: 王海南 wanghain717 wanghain717

Abstract: Intensity correlation coefficient which considered fuzzy character of image has been introduced into image registration. A new similarity measure which combines spatial information of image is developed. Experiment results show that similarity measure based on fuzzy mutual information has less local extremums and better optimization character.

Key words: mutual information, similarity measure, fuzzy mutual information, medical image registration

摘要: 从图像配准的模糊性出发,在图像配准中考虑图像的灰度相关性,结合了图像的空间信息。在配准准则中引入图像配准的灰度相关系数,提出了基于模糊互信息的配准准则。实验表明,模糊互信息函数比互信息函数具有更少的局部极值点和更好的优化特性。

关键词: 互信息, 配准准则, 模糊互信息, 医学图像配准