一种重同步的音频水印算法

Audio Watermarking Algorithm With Self-Synchronization

本文应用通信网同步方法,用VC实现了一种快速重同步音频有意义水印盲检测算法。算法的特点如下:(1)利用时域水印技术嵌入同步信息以对抗音频在时间轴上可能受到的攻击;(2)在DCT域嵌入有意义的数字水印,在保证不可感知性和稳健性的良好平衡下实现了水印的盲检测;(3)应用BCH编码来降低水印的检测错误率。同时将多数决定原则结合在语音水印和同步码的纠错中,极大地提高了水印的稳健性。

In this paper the synchronization scheme in communication network is adopted and a fast - resynchronized meaningful blind - detection audio watermarking algorithm is realized. The feature of this algorithm is that: (1) with the technology of spatial watermarking synchronization signal is embedded into the audio signal in order to resist some possible attack along the time axis; (2) Digital watermark is embedded into the DCT domain of the audio signal. Blind detection is realized with the good balance between inaudibility and robustness; (3) BCH coding is applied in order to lower the error rate of detection. We propose a new algorithm with majority rule which has very excellent effect.