一种MPEG-2压缩域视频水印算法

Novel MPEG-2 Compression Domain-based Video Digital Watermarking Algorithm

基于MPEG-2的标准,对帧内编码帧(I帧)提出一种新的水印算法,将一幅灰度图像作为水印信息。不同于传统算法的是,不需要将水印图像嵌入到原图像中,而是提取当前I帧图像的特征,利用这个特征跟水印图像建立一个版权映射。当怀疑帧内容受到侵权或修改时,利用版权映射就可以进行版权识别。而对帧间编码帧(P帧、B帧),选择其中部分运动矢量来嵌入水印。通过实验数据可知,算法对于帧删除、帧插入和帧重组等视频攻击具有很强的有效性和鲁棒性。

This paper proposed a novel video-watermarking scheme based on MPEG-2 compression standard. For the Intra frames （Ⅰ-frarnes）, a gray image was used as watermark. The most maior difference between the traditional method and the novel scheme was it did not need embed the watermark image into the original public image. Instead, extracting the feature of the I-frarnes, an ownership map according to the feature of the Ⅰ-frarnes and the watermark image could be gained. When piracy happened, the copyright owner could reveal his right via the ownership mapping and the feature of the suspected image to extracting the watermarking, For the Inter frames （P-frarnes）, partial motion vectors to embed the watermark could be used. According to the experimental results, the proposed methods were efficient, real-time, blind and robust.