一种基于DCT的扩频音频数字水印技术

首先将二值水印图像经过降维操作转化为一维序列,再将一维序列与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过程嵌入到离散余弦变换后的系数中;提取水印时对含水印信号的音频信号也进行分段离散余弦变换,然后利用原始音频信号提取水印.实验证明了该技术的鲁棒性和不可感知性.

DCT - based Audio Digital Spreading Spectrum Watermark Technology

First, the binary watermarking image was converted into a one-dimension sequence, which was then modulated with the m-sequence, before the audio signal was finally divided into segments and DCT was performed. The watermark can be extracted by DCT on the audio signal embedded watermark. Experimental results demonstrate that the technique is robust and imperceptible.