Pixel-level Image Fusion Algorithm Based on Maximum Likelihood and Laplacian Pyramid Transformation ⋆

In this paper, we proposed a novel pixel-level image fusion approach based on maximum likelihood and Laplacian pyramid decomposition. The input images are decomposed by using the Laplacian pyramid transformation at first, then the maximum likelihood method is used to fuse the image components in the low frequency bands. The selection strategy in fusing the high frequent band involving the entropy metrics theory. The final fused image can be obtained by using the inverse reconstruction of Laplacian pyramid transformation. Experimental result show that the proposed method outperforms conventional fusion methods such as discrete wavelet transform(DWT), Laplacian pyramid decomposition method, and so on