基于混合域的音频零水印算法

提出了一种基于小波变换和倒谱变换的音频零水印算法。算法对音频数据的小波逼近系数进行倒谱变换,根据倒谱系数均值与阈值T的关系及生成的索引序列向音频作品中嵌入有意义的零水印。算法采用了重复编码和PN序列调制来增强水印检测的鲁棒性,且检测时不需要原始音频,为盲检测。实验证明,算法对常见的攻击具有较强的鲁棒性。

An audio zero-watermark algorithm based on hybrid domain

This paper presents an audio zero-watermark algorithm based on wavelet trasform and cepstrum transform. In this algorithm, the wavelet approximation coefficients of the audio data are trastormed with cepstrum, and then imbed significant zero-watermark in the audio work according to the relationship between the mean of cepstrum coefficients and the threshold T and the generated index series, The algrithm uses repeating coding and PN sequences to enhance the robustness of the watemark-detecting, and the original audio data is not used when detecting, that is blind extraction. The experiments prove this algorithm has strong robustness to common attacks.