基于形态学梯度和互信息的医学图像配准方法

基于互信息的图像配准方法,已被广泛用于医学图像的配准.但是该方法计算量较大且无法处理图像空间信息,导致运行时间较长且易陷入局部极值.为解决此问题,本研究提出了一种基于形态学梯度和互信息相结合的医学图像配准新方法,该方法充分利用图像的灰度信息和空间几何形状,可节省运行时间且有效改善传统互信息方法中的局部极值问题.实验结果表明,该方法的配准精度和运行速度明显优于传统方法.

Medical Image Registration by Combined Morphology Gradient and Mutual Information

Mutual information has been widely applied in medical image registration because of the merits of non-preprocessing and high automation. However, the computed quantity of mutual information is large, so its running time is long, and it is easy to get in local maximum with absence of the processing of space information. In order to solve the problem, a new image registration method based on morphology gradient and mutual information was proposed. The running time could be shortened by the new method and the traditional local maximum problem by using intensity information and space geometric figure could be amended. The results of experiment show that the accuracy and ruaning speed of new method is superior to the traditional mutual information method.