Font Size: a A A

Research Of Medical Image Registration Based On Genetic Algorithm And Multi-Resolution

Posted on:2015-12-20Degree:MasterType:Thesis
Country:ChinaCandidate:L LiuFull Text:PDF
GTID:2298330431983941Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As the foundation of medical image fusion and surgical navigation technology, medical image registration plays an important role in medical image processing and has a great meaning for overall improvement in medical levels. In recent years, the medical image registration is a research hotspot and has a certain application.There were some associated medical devices examples in some foreign countries. In each part of medical image registration, people have carried out extensive research, in which the registration based on mutual information had better performance and robustness.This article introduces some basics of the medical image registration, and focuses on the maximum mutual information theory, genetic algorithms, multi-resolution image decomposition based on wavelet transform.It implements image registration between two brain bones images, using normalized mutual information as similarity measure, genetic algorithm as the search strategy. Finally, It utilized wavelet pyramid decomposition of image complete a multi-resolution medical image registration.On the one hand, this paper has studied the optimization process of genetic algorithm, and adopted a three-point crossover and mutation to increase the amount of three new individual for the different characteristics in different evolution periods of the genetic algorithms. To further improve the accuracy of the image registration, the rotation angle parameter code length is longer. From the results of the experiment, the registration had high stability and accuracy.On the other hand, we use image decomposition techniques based on wavelet transform complete one layer wavelet decomposition of the original image to apply multi-resolution accelerating image registration, and achieved good results.Experimental results showed that the proposed improved genetic algorithm had improved the accuracy for medical image registration, the use of multi-resolution technology in medical image registration can reduce the registration time, and to some extent, there is a good effect.
Keywords/Search Tags:medical image registration, maximum mutual information, genetic algorithms, wavelet pyramid
PDF Full Text Request
Related items