基于压缩域水印的AVS视频认证*

为了提高压缩域视频水印的安全性，提出一种基于压缩域的AVS视频认证水印新算法。它从视频空间的三维特性出发，利用I帧分块能量关系、GOP结构信息、帧结构信息等构造特征码。在嵌入算法上，将特征码同时嵌入到中频DCT系数和运动矢量中，在认证端将重建特征值和提取的双水印进行比较。仿真实验结果表明，该算法具有较好的窜改认证能力。

Compressed-domain watermarking for AVS video authentication

In order to improve the security of video watermarking, this paper proposed a watermarking-based content authentication scheme for AVS in compressed domain. It made use of the 3-dimensional characteristics of video space, and built the structure signature from the energy relationship between block of I frames, GOP and frame structural information. Both utilized the mid-frequency DCT coefficients and motion vectors for watermark embedding, made authentication by comparison between the extracted dual watermarking and restructure of the signature. Experimental results demonstrate that the proposed approach is of better capability of tampering location.