盲音频数字水印算法

Blind Audio Digital Watermark Algorithm

为有效实现网络环境下数字产品的版权保护和信息安全,在m序列调制的量化音频水印算法的基础上,提出了一种新的盲音频数字水印算法,将二维二值图像水印降为一维序列,再将水印序列与m序列做扩频调制,最后对音频信号进行分段离散余弦变换,把经过调制的水印信号经过量化处理嵌入到变换域系数上.在水印的提取过程中不需要原始的音频信号.仿真实验结果证明了算法的不可觉察性和稳健性.

In order to realize intelligence property protection and information security of digital products in network communication, the paper proposes a novel blind audio digital watermark algorithm on the foundation of modulated with m - sequence. First, the algorithm converts the two - dimension binary watermark image into the one -dimension sequence. Then, the sequence is modulated with m -sequence. At last, the audio signal is divided into segments and DCT is performed. The modulated watermark is embedded into the coefficient of the transform domain by quantification processing. The watermark can be extracted without the original audio signal. Simulation results demonstrate the algorithm is robust and imperceptible.