Infrared and visible image fusion based on improved NSCT and NSST.

Image fusion has several critical applications, which reveals its importance. The substantial information from infrared and visible images can be combined with effective image fusion. There is ample destroyable directional information that cannot be leveraged by several image fusion methods, specifically, the images taken from a long distance, with low resolution, in the dark, during bad weather, and many more. To address these problems, we have improved the non-subsampled contourlet transform (NSCT) and non-subsampled shearlet transform (NSST) by adaptive pre-processing. This paper also describes the basic structure, working principle, limitations, and applications of NSCT and NSST. Furthermore, we have presented multiple fusion results and explained the differences. Finally, we have discussed the pros and cons and a few recommendations for improving NSCT and NSST.