基于DWT-DCT和Zernike矩的鲁棒视频水印算法

为了提供数字视频版权的有效保护,提出一种基于DWT-DCT和Zernike矩的鲁棒视频水印算法。在水印嵌入阶段,先将视频帧从RGB空间转换到YCbCr颜色空间;从YCbCr图像帧的亮度分量中选择出嵌入区并进行分块,通过伪随机法选择出多个子块;执行DWT-DCT获得前n个交流分量,并将水印嵌入到这些分量中;执行DWT-DCT反变换获得水印图像,并计算Zernike矩作为校正几何攻击的参数。在水印提取过程中,通过水印嵌入阶段的伪随机密钥来确定嵌入位置,以此进一步提高水印安全性。实验结果表明,该算法具有较高的水印隐蔽性和鲁棒性。

ROBUST VIDEO WATERMARKING ALGORITHM BASED ON DWT-DCT AND ZERNIKE MOMENTS

In order to provide effective protection of digital video,we propose a robust video watermarking algorithm based on DWT-DCT and Zernike moments.In the watermark embedding stage,the video frame was converted from RGB space to the YC bC r color space.The embedding region was selected by the luminance component of the YC bC r image frame,and then divided into blocks.A plurality of sub-blocks was selected by the pseudo random method.After That,we performed the DWT-DCT to obtain the first n AC components and embedded the watermark in these components.Finally,DWT-DCT inverse transform was used to obtain the watermark image,and the Zernike moments were calculated as parameters to correct the geometric attack.In the process of watermark extraction,the pseudo random key in the embedding stage was used to determine the embedding location,so as to further improved the watermark security.The experimental results show that the proposed algorithm has high watermark concealment and robustness.