一种新颖的音频双水印算法

A Novel Watermarking With Dual Watermarks

论文提出一种既能对音频作品进行版权保护又能对其进行内容认证的音频双水印算法。首先对音频信号进行整数小波分解,然后对低频系数运用零水印技术构造鲁棒水印,达到对音频作品进行版权保护的目的。同时对整数小波分解后的高频系数进行量化处理生成脆弱水印,达到对音频作品进行内容认证的目的。实验表明,双水印系统中的鲁棒水印对常见的音频水印攻击具有较强的鲁棒性,脆弱水印对滤波等常见的信息处理具有很强的敏感性。

This paper presents a novel dual watermarking of digital audio signal.First the robust watermark is constructed by the DWT low frequency coefficients using zero-watermark technology.Then,the fragile watermark is embedded bv quantizing amplitude of DWT high frequency coefficients of digital audio signal.Experimental results show that robust watermark is robust to many operations to digital signal,such as filter,re-sampllng and the fragile watermark is inaudible and sensitive to several operations,such as filtering,MP3 compression,re-sampling,performed on audio signal.The algorithm can not only protect digital audio signal＇s copyright but also verify its originality.