Visible and infrared image fusion using NSST and deep Boltzmann machine

This paper proposes a novel fusion method for visible and infrared images based on non-subsampled shearlet transform (NSST) and deep boltzmann machine (DBM). As a typical model in the area of deep learning, DBM has remarked superiorities over several current models in terms of the function efficiency and final results. On the other hand, NSST is a novel multi-scale geometry analysis tool, and recent experimental results show that it has not only much better feature capturing ability, but also much lower computational complexities. In this paper, NSST is responsible for decomposing the source images into a series of sub-images and reconstructing the final fused image. DBM is used for conduct the coefficients selection in the sub-images. The simulation experimental results show that the proposed method has obvious advantages over the current fusion methods.