Image Registration Methods Based On Wavelet Transformation

Image registration is a fundamental problem in image processing,which is a process to match two or more images of the same scene taken at different times,from different viewpoints,or by different sensors. It is a basic step for image mosaic,object identification,image fusion,detection of changing object,analysis of sequential images etc.Application of image registration is involved in military affairs,remote sensing,medicine and computer vision etc.Through the research of many years,image registration technique has obtained plentiful research result.This paper first makes a brief summary of the phylogeny of image registration , application of image registration to the image processing and give a image registration algorithm framework for a brief overview, then does a brief introduction about the wavelet theory . Later we introduce the image registration on principle and introduce algorithm based on the gray-scale image registration and feature-based image registration. Then in view of the shortage of the traditional edge detection measures,a method of edge detection based on the wavelet transform is presented,which makes good results and proves the superiority of the wavelet transform in the edge detection . This paper also highlights the image registration algorithm baseed on mutual information and wavelet transfom.Registartion methods based on mutual information are being used widely ,The method that relies on its own image information, is a highly efficient and stable registration algorithm, and the use of wavelet transform can improve the speed, and thus gained a registration satisfactory results.