Block based motion compensation for real time Super Resolution image restoration

Super-Resolution image restoration has the need to account for movement between frames. If there is only global motion between frames, efficient frequency or spatial domain approaches as shown in are available. But in most image sequences additional local motion is found. An efficient Super-Resolution algorithm must take this into account to produce useful results. Gives an overview on different possible techniques for estimating local motion. This paper focuses on block matching techniques, as they are widely used in video compression.