Abstract: In this paper, a modified Block Matching Algorithm(BMA) with motion correlation constraint is proposed at first. Then a novel motion estimation algorithm, which computes motion vector for each pixel by interpolating motion vectors is presented. In order to increase inter-frame prediction gain and decrease the computational complexity, a motion vector optimizing algorithm for local image region is described at last. Experimental results show that the proposed algorithm can improve the prediction performance obviously with a moderate increased complexity compared with the conventional BMA.