Image Fusion Based on NSCT and Bandelet Transform

Abstract

A novel image fusion algorithm based on non-sub sampled contour let transform(NSCT) and Bandelet transform is proposed in this paper. As two different MGA tools, NSCT can give an asymptotic optimal representation of edges and contours in image by virtue of its characteristics of good multi-resolution, shift invariance, and high directionality, and Bandelet transform can take advantage of geometrical regularity of image structure and represent sharp image transitions such as edges efficiently in image fusion. The new method, joint this two transforms to fuse image, is proposed in this paper. Firstly, two source images are decomposed into different frequency sub bands by NSCT, and then high frequency sub bands are further decomposed by Bandelet transform. Secondly, the high frequency sub bands of two images are fused with fusion rule and fusion sub bands are reconstructed using inverse Bandelet transform. Finally, fusion low and high frequency sub bands and then fused image is obtained by using inverse NSCT. Several different experiments are adopted to demonstrate that fusion results based on proposed method contain more details and smaller distortion information than any other methods does.