关键词: HEVC, 帧内预测模式, 视频水印, 纹理方向, 鲁棒性

Abstract: Robust video watermarking algorithm for HEVC based on intra-frame prediction modes of multi-partioning is proposed. In order to eliminate the intra-frame error propagation when embedding the watermark, the algorithm firstly select embedable regions and calculate the texture direction in 4×4 luminance blocks.Secondly, we propose a scheme that the 33 angular prediction modes are divided into four pattern sets, which are recorded as follows: upper horizontal, lower horizontal, upper vertical, lower vertical. Lastly, The four pattern sets are mapped to the values of the current and next to be embedded watermarks. Once the current and next to be embedded watermarks meet pattern sets, the current 33 prediction modes are truncated into the current pattern sets and the watermark is embedded. the watermark is extracted by the texture direction and four pattern sets at decoding side. Experiments show that the algorithm has low video distortion and has little effect in encoding time. In robustness, the algorithm can be well resist the re-encoded attacks.

Key words: HEVC, intra-frame prediction modes, video watermarking, texture direction, robustness