出版社：The Institute of Image Information and Television Engineers
摘要：This paper proposes a new image restoration method using local characteristics of an image in order to improve its quality. This method restores an image based on a criterion including edge information. We obtain it from the criterion of the conventional Tikhonov-Miller (T-M) regularization method, where the Laplacian operator is replaced by an edge-adaptive operator, which is one of the five operators corresponding to four edge regions with the directions 0, 45, 90, 135 and a flat region. The conjugate gradient algorithm is applied to the above criterion to iteratively restore the image. The over-smoothing effect at an edge, which often occurs in the conventional restoration methods, is prevented by the edge-adaptive operator that smoothes the edge along only one direction. The edge orientation is extracted for each pixel from a temporarily restored image by the conventional T-M method.
