基于混合域的双功能音频水印算法

摘要

A dual function of watermark scheme in hybrid domain is proposed ,which can realize both the digital audio copyright protection and content authentication . Firstly , the original digital audio is decomposed into many segments according to the size of the watermark . Secondly , the DWT is performed on each audio seg‐ment .Finally ,robust watermark is embedded in the low‐frequency component while the fragile watermark is embedded in the high frequency component .Independent blind detection was achieved for both robust and frag‐ile watermarks .Simulation experiments show that the robust watermark can be used for audio copyright pro‐tection ,because the signal has strong robustness in common audio signal processing ;while the fragile water‐mark is sensitive to common signal processing operations , and may have accurate location for tampered area .%本文提出一种基于混合域的双功能水印算法，能同时实现对数字音频的版权保护与内容认证。该算法根据水印的大小将原数字音频分段，对每一段进行多级DW T变换，并选取低频分量嵌入鲁棒水印，高频分量嵌入脆弱水印，检测时两种水印独立盲提取。仿真实验表明，此算法中的鲁棒水印信号在常见的音频信号处理中具有强鲁棒性，可用于音频版权保护；而脆弱水印信号不仅对常见的信号处理操作敏感，而且能对篡改区域进行较准确的定位。