Research Of Several Problems About Image Fusion And Restoration

Image fusion is an image processing technology that combines multiple images captured by the same or different sensors from the same scene, aiming at integrating complementary information and redundant information of multiple images to obtain better image description of the scene. The result of image fusion is more suitable for human and machine perception or further image processing tasks. Image will be often polluted in the processing of acquisition and transmission. So it is important to restore image before image fusion. Image restoration is a crucial step in image fusion. Among unsolved key problems in the field of image fusion, we mainly study the image restoration in the image fusion preprocessing, infrared and visible image fusion technology, and multi-focus image fusion technology. The main contents are listed as follows:1. A novel anti-nuclear radiation image restoration algorithm is proposed. Firstly, the damaged regions are detected. Then, the damaged regions are repaired with fast digital image inpainting technology. The experimental results demonstrate that the proposed method can remove the damaged regions in the source image effectively and further enhance the performances of the camera in strong radiation environment.2. A novel fusion algorithm for infrared and visible images based on visual saliency is proposed. The method considers both the global and local characteristics of the source images. Under the image fusion framework with non-subsampled Contourlet transform, the global visual saliency feature and local variance are used to guide the low frequency coefficients and each high frequency direction coefficients fusion, respectively. The proposed algorithm emphasizes the global “interesting” object in the fused image and the visual background is more natural.3. A novel superpixel based multi-focus image fusion algorithm for raw images of single-sensor color imaging devices which incorporates the Bayer pattern is proposed. This performs fusion on the raw Bayer pattern images before color demosaicing. With the proposed fusion algorithm, the interpolation errors and amount of calculation introduced by the repeated demosaicing operation on multi-source images can also be greatly reduced. In detail, the fusion operator is performed on superpixels, and a clarity measurement of Bayer pattern image is defined to judge the focus-level of raw Bayer pattern images. By comparing the clarities of superpixels, a weight map is constructed and the guided filter is utilized to refine the weight map. The raw images are merged with refined weight map to get the fused Bayer pattern image, which is restored by the demosaicing algorithm to get the full resolution color image. Experimental results demonstrate that the proposed algorithm can obtain better fusion effect.