Detection And Restroation Of The Artifacts In Digital Videos

In this paper, we study two kinds of artifact issues in digital videos: blotches removal and mosaic detection, using image smoothing, image segmentation, image recognition, image restoration and other related digital image processing technologies.Firstly, according to the study of the characteristics of the blotches in digital videos, blotches detection and restoration model are designed and implemented finally. On the basis of the motion compensation method, this model detects blotches with ROD (Rank Ordered Differences) detector, and then restores the blotches by means of MMF (Multilevel Median Filter) interpolation method. Secondly, the features and the cause of the mosaic in digital videos are analyzed, and some related image processing technologies, such as edge detection, template matching and binarization, are employed to design and implement the sliding window mosaic detection algorithm. The experimental results further verify the feasibility and reliability of the algorithm.