一种改进的DCT域数字音频水印算法

针对现有数字音频水印算法不能同时满足隐蔽性与鲁棒性的问题，在添加水印之前，先将水印的二值图像与随机密钥进行加密操作，再对载体进行三级小波变化和DCT变换，最后对比加密后的水印信息与载体信号，得到密钥值，并将密钥嵌入到载体音频的DCT中频系数中。实验结果表明，该算法能在隐蔽性和鲁棒性之间达到一个较好的平衡。

An Improved Digital Audio Watermarking Algorithm Based on DCT

For the existing digital audio watermarking algorithm can＇t satisfy both hidden and robustness, before add a watermark, use a random key to encrypt the binary image of watermark, and then make three level wavelet transform and DCT transform on carrier, finally, compare the watermark information encrypted with the carrier signal to get a key, and embedded the key into intermediate frequency coefficient of carrier audio. Experimental result shows that the algorithm can achieve a good balance between the hidden and robustness.