Multi-modal Medical Image Registration Based on Gradient of Mutual Information and Morphological Haar Wavelet

摘要

In this paper, we proposed multi-modal medical image registration method, which introduces mathematical morphology into image processing, and makes full use of the non-linear multi-resolution of the morphology Haar wavelet; the image pyramid is constructed according to the different resolution, which can enhance the speed of the registration. Genetic algorithm (GA) is used to accurately obtain the optimal parameters. In addition, we use the gradient of mutual information to measure the similarity between the two images. Introducing the gradient information into the mutual information makes the registration algorithm more robust and more accurate. The experiment shows that this algorithm can do more effectively and more accurately work than the algorithm based on Haar wavelet.