摘要
提出一种改进的基于运动矢量的视频水印方案,该算法有下述优点:①根据水印嵌入容量自适应调整嵌入策略,以满足不可感知性要求;②水印嵌入对差分编码运动矢量造成的误差累积减小;③与MPEG＿2视频编码语义紧密结合。实验证明所提出的视频水印算法具有较低的计算复杂度,对视频码率转换具有高的鲁棒性。
An improved video watermarking scheme employing motion vectors is proposed. Compared with similar schemes proposed by Jordan, ZHANG and Dai, respectively, the new one described herein has three merits: (① Embedding )strategy is adaptively adopted according to the requirement of watermarking capacity to meet the need of imperceptibility; ② The accumulation of differential coding error of motion vectors is reduced;③ Compatible with the syntax of MPEG_2 coding stream. Experimental results have demonstrated that the proposed scheme is of low computational complexity and of high robustness against rate conversion.
出处
《中山大学学报（自然科学版）》
CAS
CSCD
北大核心
2004年第A02期117-119,共3页
Acta Scientiarum Naturalium Universitatis Sunyatseni
基金
江苏省自然科学基金资助项目(BK2004421
BK2004132)
