Low complexity video encoding using new motion vector skip mode coding

In this paper, we propose a new low complexity video encoding method targeted specially for power-constrained mobile devices. Since the motion estimation takes most of computation and thus battery consumption, the proposed encoding method restricts the possible block size of motion estimation to 16×16 and 8×8. It curtails considerable amount of computation off the encoder, but brings non-acceptable degradation in coding efficiency. The problem is overcome by the proposed new motion vector skip coding whose key is to best utilize the spatial correlation between motion vectors of adjacent 8x8 blocks by effective motion vector coding. The experimental results with several well-known test video sequences verify that far better performance of the proposed method is obtained against the H.264/AVC baseline encoder of similar configuration.