Seamless stitching of images based on a Haar wavelet 2D integration method

A method for seamless stitching of images with photometric inconsistencies in the overlapping region is presented. The method is based on generating a set of stitched gradients from the gradients of the input images and then reconstructing the mosaic image using the Haar wavelet integration technique of. This O(N) reconstruction technique is based on obtaining the Haar wavelet decomposition of the mosaic image directly from the stitched gradients and then using Haar synthesis to obtain the mosaic image. The Haar synthesis step includes a Poisson smoother at each resolution level leading to results without visual artifacts despite the non conservative nature of the stitched gradient field. Experimental results illustrate the method and show that it leads to seamless mosaic images despite intensity differences in the overlap region of the input images.