一种基于视频码流的视频水印方案

Video Watermarking based on Video Code Stream

为了解决视频数据的版权问题,提出了一种基于视频码流的视频水印方案。该方案首先利用秘密信息分存方案将水印信息分存为多份待嵌水印;然后,利用FFMPEG定位视频流中残差和运动向量差值的可嵌入二进制编码比特;最终,利用水印数据修改可嵌位置的比特完成水印水印信息的嵌入。理论分析与实验结果均表明,与传统方案相比,该方案具有视频体积零膨胀、对视频质量影响小的优点,且嵌入速度达到了实时。

In order to protect the copyright of video data,the video watermarking scheme based on video stream is proposed.First of all,the secret information sharing scheme is used to divide the watermark information into many shares,then those bits that can be changed to embed the watermark in the residual differences and motion vector differences are accurately located via the ffmpeg open code,and finally,the embedded bits of watermark is modified by using watermark data,thus completing the embedding of watermark information.Theoretical analysis and experimental results indicate that compared with the traditional scheme,the proposed scheme has the advantages involving zero expansion of video volume,small impact on video quality,and real-time embedding speed.