Multi-filtering based edge preserving image fusion technique

Image fusion is the process of combining information of several images into a single image. Motivated by the efficiency of multi-filtering in highlighting the image details, a new image fusion algorithm is developed in this paper. Each image is filtered along rows and columns to generate four sub-images. These images contain features along rows and columns. Each filtered image is given to the Laplacian of Gaussian (LoG) operation for obtaining weights for fusion. Output of LoG is thresholded and edge preserving filtering is applied to get two sets of weights. Sub-images are blurred, sharpened and scaled with these weights. Finally, all scaled sub-images are added to get fused image. The proposed algorithm is evaluated both qualitatively (by visual inspection) and quantitatively (using various fusion metrics) with existing recent algorithms.