Research On SAR Image And Multi Spectral Image Fusion Algorithm Based On Multi Resolution Analysis

With the rapid development of sensor technology,people can get image information in a variety of ways now.But it is very difficult to have a comprehensive presentation on the target by a single type of image,so the image fusion processing technology has emerged.The main idea of image fusion algorithm is to use some certain process to fuse the images which are taken by different image sensors at the same place.The new image is more suitable for human visual perception,or to meet the needs of segmentation,feature extraction and target recognition in image processing.This paper focuses on the multi source image fusion algorithm based on non-subsampled contourlet transform,and through combining fuzzy cluster analysis(FCM)and HIS color model transform with non-subsampled contourlet transform to fuse synthetic aperture radar image(SAR image)and multi spectral image(MS image).Under this framework,two different fusion methods are proposed in this paper.The first fusion method is based on multi windows,which will adopt different window fusion rules in the low frequency coefficient fusion according to different regions,so as to obtain more information of the source image,at the same time,when the high frequency coefficient is fused,the regional energy block is adopted to make sure the details of the image can be preserved to the maximum extent.The second fusion method is a feature level fusion method based on regional characteristics.This algorithm will fuse the low frequency sub-band coefficients based on a new weighted average method based on a variety of regional characteristics.In the high frequency sub-band coefficients fusion will be based on the correlation coefficient of the corresponding coefficients to choose different fusion methods,which not only can reduce the amount of computation,but also preserve image details.