Infrared and visible image fusion method based on fractional-order differentiation

In order to solve the problems of information loss in image fusion, an infrared and visible image fusion method based on fractional-order differentiation is proposed. Firstly, the multi-scale transform is used to decompose the source images into low frequency and high frequency subbands, and the low frequency subbands are further decomposed into low frequency basic subbands and low frequency detail subbands by two-scale decomposition. Secondly, for the low frequency base subbands, the weighted sum of energy ratio and standard deviation ratio is used to construct the judgment value which is used to fuse low frequency base subbands. For low frequency detail subbands and high frequency subbands, the fractional-order differentiation is introduced, and the fusion rule of maximum fractional-order sum of modified laplacian is adopted. Finally, the fused low-frequency basic subband and low-frequency detail subband are transformed by two-scale inverse transformation to obtain the fused low-frequency subband. the multi-scale inverse transformation is performed to the fused low frequency subband and high frequency subbands to obtain the fused image. Three groups of infrared and visible images are selected to verify the effectiveness of the proposed algorithm. From the subjective assessments, the proposed method highlights the infrared target well, retains the details of the visible image and texture details, and achieves a good visual effect. From the objective assessments, the entropy, standard deviation, spatial frequency and mean gradient of the fusion method in this paper are higher than the other five methods.