Abstract
Image processing using local patches has become very popular and was shown to be highly effective. We propose a novel approach based on reordering of its patches. For the given corrupted image, first, we extract all the patches with overlaps. Then we will order them in such a way that they are chained in the shortest possible path. The order which we obtained will be applied to the corrupted image, implies a permutation of the image pixels to what should be a regular signal. This will help us to achieve good recovery of the clean image by applying relatively simple one-dimensional (1D) smoothing operations to the reordered set of pixels. From this, we have shown that using a carefully designed permutation matrices and simple and intuitive 1D operations such as linear filtering and interpolation, the proposed scheme can be used for image denoising and inpainting, where it achieves high quality results.
