摘要
本文提出一种在原始视频DCT(离散余弦变换)与SVD(奇异值分解)域自适应嵌入水印的算法。算法首先将视频流分割成一个个场景,场景中的视频图像被变换到DCT-SVD域中,水印量化嵌入在最大奇异上,实现了水印盲检测。同时实验证明该算法在满足透明性要求下也满足一定的鲁棒性要求,能够抵抗低通滤波、中值滤波、椒盐噪声、高斯噪声、H.264压缩攻击以及统计攻击和帧剪切等攻击。
In this paper, a blind adaptive video watermarking algorithm based on DCT （ discrete cosine transformation） and SVD （singular value decomposition） is proposed. The algorithm first partitions the video stream into several shots, And both the frames of the shots are transformed into DCT-SVD domain. the watermarking sequence is embedded into the largest singular value. Experiments showed the new algorithm meet the transparency requirements at the same time meet the requirements of a certain robustness, resistance to low-pass filtering, median filter, noise and H.264 compression, anti Collusion, Cut-and-splice attacks.
出处
《微计算机信息》
2009年第33期62-63,86,共3页
Control & Automation
