针对数字视频版权保护问题,利用人类视觉系统(HVS)特性基于MPEG-2标准提出一种盲水印算法。首先选定为B帧亮度最高的3个8×8 DCT分块作为水印嵌入位置,然后利用分块位置和低频分量增长趋势实现水印的嵌入,最后通过仿真实验验证了算法的不可见性与鲁棒性。
In view of the copyright protection of digital videos, a blind video watermarking algorithm based on MPEG - 2 standard has been proposed by using human visual system (HVS) characteristics. Firstly, three 8 ~ 8 DCT blocks of the brightest B frames are se- lected as the watermark embedding positions, then the watermarking information is presented through block position and the growth trend of low frequency component, and finally the simulation experiments verify the invisibility and robustness of the algorithm.