Infrared and Visible Image Fusion Based on Innovation Feature Simultaneous Decomposition

In the field of image fusion, for the problem of large difference between infrared and visible image, these two are decomposed by joint sparse representation and get the common feature and innovation feature; Furthermore, a new fusion method based on innovation feature simultaneous decomposition is proposed to solve the problem of feature rich in infrared and visible light images. Based on the joint sparse representation model, the multi- source image data is combined into a new signal at the corresponding position, and the innovation feature of the two is decomposed into the same atom by Simultaneous Orthogonal Matching Pursuit. Experimental results and analysis show that the proposed method is superior to the popular sparse representation method, the simultaneous orthogonal matching pursuit method and the traditional joint sparse representation method, and the subjective visual evaluation and the objective index evaluation is both superior. The relevant research can be extended to the specific feature rich remote sensing image or medical image fusion field.