Photonic, Technical University of Denmark
Computer & Electrical Engineering, Babol Noshiravani University of Technology
Image restoration is an essential preprocessing step for many image analysis applications. In any image restoration techniques, keeping structure of the image unchanged is very important. Such structure in an image often corresponds to the region discontinuities and edges. The techniques based on partial differential equations, such as the heat equations, are receiving considerable attention in image restoration. In this paper, a new algorithm is proposed for image restoration using partial differential equations applied on neighbors longer than one pixel. To better preserve edges in this technique, slant edges are considered in addition to vertical and horizontal edges. A number of experiments have been preformed to evaluate performance of the proposed method and to compare its performance with the existing algorithms, wiener and median filters. The results express the considerable superiority of the proposed methods.