Fusion of Infrared and Visible Images Based on Infrared Object Extraction

Abstract

The ideal fused results of infrared and visible images, should contain the important infrared objects, and preserve the visible textural detail information as much as possible. The fused images are more consistent with human visual perception effect. For this purpose, a novel infrared and visible image fusion framework is proposed. Under the guidance of the model, the source images are decomposed into largescale edge, small-scale textural detail and coarse-scale base level information. Among which, the large-scale edge information contains the main infrared features, on this basis, the infrared image is further segmented into the object, transition and background regions by OTSU multi-threshold segmentation algorithm. In the end, the fused weights for the decomposed sub-information are determined by the segmented results, so that, the infrared object information can be effectively injected into the fused image, and the important visible textural detail information can be preserved as much as possible in the fused image. Experimental results show that, the proposed method can not only highlight the infrared objects, but also preserve the visual information in the visible image as much as possible. The fused results are superior to the commonly used representative fusion methods, both in subjective perception and objective evaluation.