Restoration of scanned photographic images

In this paper, we address the problem of restoring photographic images degraded by motion blur and film-grain noise. Based on the one-dimensional particle filter, a new approach is proposed for restoration under space-invariant as well as space-variant blurring conditions. The method works by propagating the samples of the probability distribution through an appropriate state model. The weights of the samples are computed using the observation model and the degraded image. The samples and their corresponding weights are used to estimate the original image. In order to verify and validate the proposed approach, the method is tested on several images, both synthetic and real. © 2005 Elsevier B.V. All rights reserved.