Multi-modal Image Fusion Algorithm based on Variable Parameter Fractional Difference Enhancement

Abstract Multi-modal image fusion can more accurately describe the features of a scene than a single image. Because of the different imaging mechanisms, the difference between multi-modal images is great, which leads to poor contrast of the fused images. Therefore, a simple and effective spatial domain fusion algorithm based on variable parameter fractional difference enhancement is proposed. Based on the characteristics of fractional difference enhancement, a variable parameter fractional difference is introduced, the multi-modal images are repeatedly enhanced, and multiple enhanced images are obtained. A correlation coefficient is applied to constrain the number of enhancement cycles. In addition, an energy contrast is used to extract the contrast features of the image, and the tangent function is simultaneously used to obtain the fusion weight to attain multiple contrast-enhanced initialization fusion images. Finally, the weighted average is applied to obtain the final fused image. Experimental results demonstrate that the proposed fusion algorithm can effectively preserve the contrast features between images and improve the quality of fused images.