Multi-directional detection of scratches in digitized images

Line scratches are common defects in old archived videos. Many printed images suffer of a similar problem, in most cases due to improper handling or inaccurate preservation of the support. Once an image is digitized, its defects become part of that image. Many state-of-the-art papers deal with long, thin, vertical lines in old movie frames, by exploiting both spatial and temporal information. In this paper we present a method that consistently modifies and extends our previous proposed algorithm, to detect line scratches in digitized still images. Our method is able to detect scratches regardless of their orientation, and to draw their contour by labeling the pixels belonging to them.