Digital video watermarking based on intra prediction modes for audio video coding standard

Digital video watermarking provides means for carrying information targeted for synchronization, error resilience or copyright protection. However, it is difficult to get a good trade-off between the embedding capacity, imperceptibility and efficiency. In this paper, a novel digital video watermarking algorithm based on intra prediction modes of AVS (audio video coding standard) is proposed. The algorithm hides one bit in each qualified intra 8 × 8 luma block by modifying intra 8 × 8 prediction modes based on the mapping rules between the modes and hidden bits. The specific positions of the host 8 × 8 blocks are determined by the features of every block and a position template indicated by the key. Watermark information can be retrieved by decoding the intra prediction modes from bitstream, requiring neither original media nor complete video decoding. Experimental results show that the proposed algorithm has little impact on video quality and video stream features. A comparatively high embedding rate is obtained with little impact on bit rate.