Detection and removal of anomalies in digitized animation film

In this article, the problem of detecting and removing on the development of practical and efficient algorithms to be anomalies in digitized animation film is addressed. The impetus of incorporated directly into the animation production system. this article comes from the motion picture industry, where several There has been a good amount of work in the literature dealing studios are rereleasing vintage film to the public which are often acwith the removal of artifacts in images and video. What defines companied by visual degradation in the film color, grain, and overall an artifact depends on the application. For example, owing to the quality. These degradations, or anomalies, can be seen as unwanted quantization step in image and video compression, blocking and visual artifacts that are usually larger than one pixel in size and appear ‘‘mosquito’’ artifacts appear, which need to be removed [5]. in only one frame. One of the novelties of this article is that the restoraDuring transmission of compressed images and video, packets tion of animation film, as opposed to ‘‘real-world’’ image sequences, is investigated. It often contains additional artifacts and raises other are lost, and therefore parts of the reconstructed frame are mississues not found in most sequences. These artifacts are first defined, ing; this error needs to be concealed [6] . and the appropriate steps for their detection and removal are deIn film, typical artifacts are due to dirt and sparkle, as seen scribed. Experiments with the proposed algorithm were performed in the bright and dark flashes in a frame. These artifacts are using scenes from the animation film Fantasia, courtesy of Walt Disney caused by either accumulated dirt on the film or film abrasion. Feature Animations, and are discussed in detail. Although our algoData can also be found altered or missing owing to scratches rithm was developed for the removal of artifacts on animation film, and fingerprints on the cellulose surface. The earliest automatic aspects of it can be applied to nonanimated film. q 1998 John Wiley & detection of dirt and sparkle was done in 1985 by Storey [7]. Sons, Inc. Int J Imaging Syst Technol, 9, 283–293, 1998 The disadvantage of his approach, however, was the lack of any motion compensation algorithms.