Infrared and Visible Image Fusion Based on Side Window Filtering

Integrating the complementary information of infrared and visible images into a single image is an important and challenging task in the field of image fusion. The commonly used feature-level fusion methods usually suffer from rising artifacts and noise in the fused results. To overcome these drawbacks, in this paper, a novel feature level fusion algorithm is proposed by designing an edge preserving filter and using efficient feature segmentation algorithms. First, a novel adaptive brightness enhancement algorithm is proposed to solve the problems of low contrast and insufficient illumination in visible light images. In addition, the noise problem of infrared images is dealt with by the proposed noise suppression algorithm. Then, an efficient image segmentation algorithm is employed to generate two rough weight maps. Next, to minimize the effect of artifacts and noise, a edge-preserving filter bank is designed to refine the obtained rough weight maps. In addition, the top-hat transformation is used in order to minimize the influence of background noise on the thermal radiation information in the weight map. Finally, the weight-based pyramid fusion strategy is used to obtain fusion results. Experiments demonstrate the significant improvement of fusion quality brought about by the proposed method.