For the copyright protection application of digital video, this paper proposed a video watermarking algorithm based on lifting wavelet transform (LWT) and discrete cosine transform (DCT). Firstly, the algorithm preprocessed the watermarking by using chaotic encryption and Arnold scrambling. Secondly, it selected randomly r color frames with a key from a video, and subdivided each channel of selected frames to non-overlapping blocks of size 8×8 and performed 1 -level LWT for each selected blocks. Finally, it applied DCT for each low-frequency sub-band of selected blocks and embedded the watermarking in the DC component by using the texture and motion characteristics of selected frame to determine adaptively the quantization step of dither modulation and could be extracted without the original video. Experimental results show that the proposed algorithm is simple and has good transparency and robustness. Compared with other algorithm, it has better performance.%针对数字视频的版权保护应用,提出了一种混合提升小波变换和DCT的视频水印算法.该算法先对水印进行混沌加密和Arnold置乱处理,借助密钥选取r帧彩色视频并将每帧视频的每一分量进行互不重叠的8×8分块,对选取的分块进行1级提升小波变换,并对低频子带进行DCT变换,以视频帧的纹理和运动特性自适应地确定量化步长的抖动调制方式嵌入水印,水印提取时无须原始视频的参与.实验表明,该算法实现简单,具有良好的透明性和鲁棒性,与其他算法相比,该算法具有更好的性能.
展开▼
