Infrared and Visible Image Fusion using Multi-Scale Decomposition and Visual Saliency Map

Abstract

In this paper, a novel infrared (IR) and visible (VIS) image fusion algorithm is proposed by using a joint-histogram weighted median filter (JH-WMF) and a mean filter (MF). The proposed method is based on a multi-scale decomposition (MSD) of an image into a base layer, a texture layer, and an edge layer, which is helpful to suppress the artifacts. Furthermore, a novel saliency detection algorithm is proposed for base layers fusion by combining JH-WMF and MF, which is more effective to reduce the loss of contrast. Experiments show that the proposed approach achieves better performance than other methods, in terms of subjective visual effect and objective assessment.