结合人类听觉系统,提出了一种基于倒谱变换的自适应音频水印算法,充分利用复倒谱变换的性质,将原始音频信号分成若干帧,每帧实施复倒谱变换后,在对应位置按照一定的方法嵌入水印信号。水印的提取不需要原始音频信号,是一种盲水印算法。实验结果表明,嵌入后的水印不仅具有很好的不可感知性,而且对添加噪声、重新采样、低通滤波和重新量化等攻击也具有很好的鲁棒性。
The paper proposed an adaptive audio watermark algorithm based on cepstrum transform. After dividing original audio signal into several segments, every segment was transformed with complex eepstrum. Embedded watermark using masking of human auditory system (HAS) and the characteristics of complex cepstrum transform. The watermark could be extracted without the original digital audio signal. Experimental results show that embedded watermark is imperceptibility and robust to many attacks, such as noise adding, re-sampling, low pass filtering and re-quantization and so on.