Infrared and Visible Image Fusion Algorithm Based on Threshold Segmentation

Abstract

For advancing the fusion algorithm quality of infrared and visible images, an image fusion algorithm of infrared and visible images was presented by threshold segmentation. The visible image was turned to HSV color space, and the V component of visible image and the infrared image were decomposed multi-resolution by Laplacian pyramid transform. Using the least squares principle, low-frequency (LF) coefficients of the infrared image were segmented by Otsu threshold segmentation to enhance infrared thermal targets in the fused image. Binary image, which was generated by Otsu threshold segmentation from LF coefficients of infrared image, was an important fusion rule. Multi-scale coefficients of Laplacian pyramid transform were fused by rules of region mutual information, matching degree, and region energy. Lastly, the fusion image reconfiguration was realized through Laplacian pyramid inverse transform. The experimental result shows that the image fusion algorithm presented by this paper retains clear scenes of the visible image adequately, meanwhile highlights the thermal targets of the infrared image, which enriches information and object characteristics in the fusion image.