Enhanced infrared and visible image fusion using a fast guided filter and an improved visual saliency map

Abstract

The fusion of infrared (IR) and visible images aims to generate a composite image that is both highly informative and visually optimized for human perception or computer vision applications. This paper presents an innovative fusion approach designed to enhance the effectiveness of this process. The method begins by utilizing the Fast Guided Filter (FGF) to decompose input images into base and detail components. An improved visual saliency map technique is introduced to efficiently fuse the base components, while a local energy-based fusion strategy is applied to the detail components. Experimental evaluations demonstrate that the proposed approach outperforms conventional methods, delivering superior results in both subjective and objective assessments. Keywords: Fast guided filter (FGF), visual saliency map (VSM), image fusion (IF)