基于改进的互信息结合边缘互方差的医学图像配准方法研究

研究了灰度多模医学图像配准中广泛使用的最大互信息方法,在此基础上,又研究了边缘图像的互方差配准原理,并将其引入到原始图像的互信息中,综合互信息和边缘互方差对图像配准的目标函数作了改进,定义了一个新的测度函数.通过使用Powell算法和模拟退火混合优化算法对目标函数进行寻优,得到了配准变换参数,最终对图像实现了配准.这种方法既利用了原始图像的灰度互信息,又利用了边缘图像在二维水平面上像素间的关系,起到了补偿互信息缺陷的作用,提高了配准的精度.通过Matlab仿真实验证明这是一种有效的医学图像配准方法.

An Improved Method for Medical Image Registration Based on Mutual Information Combined with Edge Correlation Deviation

This article studies the maximum mutual information method in application of gray-scale multi-modal medical image registration.On this basis,it also studies the registration principle of the edge correlation deviation which is brought into the mutual information of original images.This paper improves the goal function of registration at basis of mutual information.It integrates the edge correlation deviation and defines a new measure function.Registration transformation parameters are obtained by using Powe...