基于互信息图像配准中的局部极值问题研究  

Study on local extremum of object function in mutual information-based image registration

作　　者：叶波[1] 陆雪松[1] 张素[1] 陈亚珠[1] 

机构地区：[1]上海交通大学生物医学仪器研究所,上海200240

出　　处：《计算机工程与应用》2007年第6期58-61,共4页Computer Engineering and Applications

基　　金：国家重点基础研究发展规划(973)(the National Grand Fundamental Research 973 Program of China under Grant No.2003CB716103) 。

摘　　要：基于互信息的图像配准方法具有自动化程度高、配准精度高等优点,近年来在医学图像配准中得到广泛应用。但是当变换后像素坐标位于非采样网格点时,插值算法有时会使目标函数产生局部极值,使得最优化搜索终止于局部极值,得到错误的配准结果。分析了两种常见的插值算法和产生局部极值的原因,在此基础上提出一种新的插值算法。实验结果表明,该算法有效地抑制了基于互信息的目标函数的局部极值问题,使目标函数更加平滑。Mutual information-based image registration has been widely exploited due to its high-level automation and accuracy, especially in medical image registration.However,the object function suffers from the local extremum induced by interpolation, which obstructs optimization algorithms and leads to the wrong result of registration.The mechanisms how two common used algorithms work and the reason why the artifacts occurred are analyzed and a new interpolation algorithm is proposed in this paper.Experimental results indicate that the new interpolation algorithm can effectively avoid the local extremum,and smooth the object function through mutual information.

关 键 词：图像配准 互信息 局部极值 插值 

分 类 号：R319[医药卫生—基础医学] TP751[自动化与计算机技术—检测技术与自动化装置]