Adaptive spatial prediction in intra coding

In this paper, a novel spatial prediction method is proposed to improve prediction accuracy and capture the dynamics of different video contents. The method adaptively generates optimized spatial prediction modes according to the local feature of coded image. The optimization is conducted by analytical calculations to minimize the energy of prediction residue. Experimental results show that the method is especially effective for high resolution sequences and high bit rate coding scenarios. Compared to H.264, the proposed method improves the coding efficiency by 3.90% in average for all sequences and 4.64% in average for high resolution sequences (1080p and 2560×1600).