摘要
提出了一种自适应映射能量(AMEW)视频水印算法,将水印嵌入到压缩域中,从压缩码流中提取水印,提取过程不需要原始视频,是一种新型的盲视频水印算法.在该算法中,作者选择一定的整数DCT量化系数构造嵌入空间,然后利用能量映射函数将水印嵌入其中.在嵌入过程中,作者利用三个模块(最大匹配串模块,最小更改串模块和最小丢弃串模块)调整嵌入后的码流来降低对编码效率(通常我们用视频质量和压缩率来衡量编码效率)的影响,实现最优嵌入;重同步策略和水印编码是用来保证水印的鲁棒性.实验结果证明水印对于重编码、帧删除、变帧率及锐化等操作具有一定的鲁棒性,几乎对编码效率没有影响.除此以外,在同码率条件下,与XDEW算法相比,该算法还具有高水印负载和低计算复杂性的优势.在该文末尾,作者将阐述该算法的理论根据并分析其鲁棒性.
The paper presents a novel video watermarking scheme called Adaptive Mapping Energy video Watermarking（AMEW）for copyright protection. The watermark is embedded in the compressed domain and extracted directly from the bitstream without original video. In the proposed scheme, authors select a part of Integer DCT quantized coefficients to construct the embedded space and embed the watermark into it by using Energy Mapping Function（EMF）. During the process of embedding, such three modules as Longest Common Sequence（LCS）module, Least Modification（LM）module and Least Casting（LC）module are used to adjust the watermarked bitstream to attain the least influence of coding efficiency（high video fidelity and compression rate） ; re-synchronization strategy and watermark coding are used to guarantee the robustness of watermarking. The experimental results indicate that the scheme has strong robustness to the attacks such as re-encoding, frame dropping, frame rate changing and sharpening. The influence of the coding efficiency is almost unnoticeable. Besides, high watermark payload and low time complexity are advantages of the scheme compared with XDEW at the same bit rate. At the end of the paper, authors analyze the theory base and robustness of AMEW.
出处
《计算机学报》
EI
CSCD
北大核心
2005年第11期1843-1852,共10页
Chinese Journal of Computers
基金
国家自然科学基金(60333020,60472043)资助
