Research on Image Fusion Based on the Nonsubsampled Contourlet Transform

Abstract

As a new image multi-scale geometric analysis tool, the nonsubsampled contourlet transform (NSCT) has many advantages such as multi-scale, localization and multi-direction, and can effectively capture the geometric information of images. Therefore, when the NSCT is introduced to image fusion, the characteristics of original images can be taken better and more information for fusion can be obtained. In addition, due to the elimination of the downsampler and upsampler during the decomposition and the reconstruction of the image, the NSCT has such characteristics as the shift-invariance and the same size between each subband image and the original image, which will effectively reduce the effects of the mis-registration on the fused results and can make it easy to find the corresponding relationship of each subband image. Therefore, the NSCT is applied to image fusion and a novel algorithm for fusion of multi-sensor images based on the NSCT is proposed in this paper. The algorithm has been used to merge several multi-focus images. The experimental results indicate that the proposed approach can avoid the introduction of the artifacts or the high frequency noise and can also significantly outperform the traditional wavelet-transform-based image fusion method in terms of both visual quality and objective evaluation criteria, especially in situation where the source images are not perfectly registered.