针对压缩域视频水印安全性与误差积累问题,提出一种基于AVS压缩域的水印新算法。针对误差积累问题,提出两种控制方法,最大限度地降低了误差积累。此外,对嵌入前的水印进行了混沌处理,提高了水印的安全性。仿真实验结果表明,误差积累得到很好的控制,满足盲检性和不可见性。
According to the security of watermarking and the problem of error propagation, proposed a novel for AVS. Two compensation methods to overcome the error propagation after watermark embedding into motion vectors, which would greatly decrease the error propagation. Moreover, the chaotic encryption of watermark greatly improved the security of watermarking. Experimental results on several test sequences demonstrate that the proposed approach controls the error propagation well and can meet the requirements of blind detection and imperceptibility.