首页 | 本学科首页   官方微博 | 高级检索  
     

基于小波金字塔和轮廓特征的医学图像配准
引用本文:张石,唐敏,董建威.基于小波金字塔和轮廓特征的医学图像配准[J].计算机仿真,2008,25(5):205-209.
作者姓名:张石  唐敏  董建威
作者单位:东北大学信息科学与工程学院,辽宁,沈阳,110004
摘    要:提出一种快速而精确的自动医学图像配准算法.先提取图像的轮廓,根据力矩主轴法计算初步配准变换参数的值,实现参考图像和待配准图像的初步配准;然后在此基础上,利用小波分解图像,提取图像的边缘特征点,采用小波金字塔分层搜索的策略在一个相对较小的范围对变换参数进行迭代搜索,最终实现图像的精确配准.不但提高了配准的精度,而且可减少精配准过程中所需的时间,有效地降低了配准过程所需的计算量.由实验结果可以看出,配准结果精确,实现速度快,因此是一种实用的多模态配准算法.

关 键 词:医学图像配准  力矩主轴法  小波金字塔  交叉相关
文章编号:1006-9348(2008)05-0205-05
修稿时间:2007年4月28日

Automatic Registration of Medical Image Based on Wavelet Pyramid and Contour Feature
ZHANG Shi,TANG Min,DONG Jian-wei.Automatic Registration of Medical Image Based on Wavelet Pyramid and Contour Feature[J].Computer Simulation,2008,25(5):205-209.
Authors:ZHANG Shi  TANG Min  DONG Jian-wei
Affiliation:ZHANG Shi,TANG Min,DONG Jian-wei (College of Information Science , Engineering,Northeastern University,Shenyang Liaoning 110004,China)
Abstract:In this paper, a fast-speed and accurate automate registration method for medical image based on wavelet pyramid and contour feature is described. Firstly, using principal axes algorithm based on the image edge to realize preliminary registration. Secondly, for precise registration, using wavelet transformed pyramids to provide a detailed set of matched points quickly. After generating the image pyramids, the cross-correlation algorithm is proposed to search identical points among images on the same level. ...
Keywords:Medical image registration  Principal axes algorithm  Wavelet pyramid  Cross-correlation algorithm  
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