Multifocus, Infrared and Visible Light Image Fusion Using Non-subsampled Shearlet Transform and SUSAN Operator

Image fusion is the multipurpose and multidisciplinary field studied by most of the researchers these days. This research applies non-subsampled shearlet transform (NSST) and a human visual system operator to fuse multifocus and infrared visible light image pairs. Initially the images are decomposed in different subbands using NSST and then local energy and SUSAN operator based fusion rules are applied on these subbands. The fusion results are compared visually and using different performance measures with three existing fusion schemes. The fusion comparison of the schemes show that the proposed scheme is capable of fusing the multifocus and infrared visible image pairs. The average run time of the proposed scheme is quite low as compared to other schemes.