Automatic image registration using evolutionary algorithm

Abstract

Image registration is a process to obtain the spatial transformation of an input image to a reference image by which similarity measure is optimized between the two images. Mutual information is a similarity measure based on information theory used in the process of image registration. Mutual information compares the statistical dependency between images. Registration based on mutual information is robust and could use for a large class of mono modality and multimodality images. In this work we use mutual information as the similarity measure .There is a requirement to finding the global maxima of similarity measure, for this we use two algorithm simple genetic algorithm and share genetic algorithm and compare the result of these algorithm. In these optimization technique require several decision to made during implementation, such as encoding, selection method and evolution operator. In this work we use two selection method roulette-wheel method and tournament selection method. Result indicates that these optimization techniques can be used for efficient image registration