Recovery of motion vector based directional similarity

A new approach to motion vector recovery in video compression is presented in this paper. The approach is based on the assumption that a group of macroblocks (MBs) which belongs to the same object and resides in the same region likely to move in the same direction. Hence the proposed technique estimates a lost motion vector (MV) from a selected group of candidate MVs. A group of candidate MVs is chosen from the corresponding MVs of surrounding MBs on the left, top-left, top, top-right, right, bottom-right, bottom and bottom-left based on their directional similarity. The experimental results for several test video sequences are compared with conventional error concealment methods and higher performance is achieved in both objective peak signal-to-noise ratio (PSNR) measurements and subjective visual quality.