An improved motion vector competition scheme for accurate motion vector prediction

Abstract

H.264/AVC includes lots of techniques for video compression efficiency and especially, the motion estimation technique achieves considerable coding efficiency by reducing the temporal redundancy between frames in video sequences. Since accurate motion vector prediction can reduce more redundancy, motion vector prediction is one of the most important parts in motion estimation. In this thesis, an improved motion vector competition scheme is proposed for accurate motion vector prediction. The proposed method consists of selection of candidate predictors among motion vectors of the neighboring blocks of the current block and the co-located motion vector in the previous frame, scaling of the candidate predictors for multiple reference frames, and motion vector competition according to a coding mode. In addition, it introduces a method to avoid an overhead of index bits that need to determine the type of the predicted motion vector. Experimental results show that the proposed method improves the coding performance for various video sequences.