为了提高传统水印方案的透明性,该文对基于离散小波变换(DWT)和离散余弦变换(DCT)的音频零水印进行了研究。该文将音频分成等长帧。对每帧进行DWT提取出近似分量。对近似分量进行DCT后求均值。根据相邻均值间的关系嵌入水印。仿真实验表明该算法对于MP3压缩、重量化、重采样、低通滤波、降噪等常见音频信号处理攻击具有很强的鲁棒性且效率较高。
In order to improve the imperceptibility of traditional watermarking scheme, a novel audio zero-watermarking scheme based on discrete cosine transform and discrete wavelet transform (DCT- DWT) is proposed. The audio is split into equal frames. Approximate components are extracted using the DWT for each frame. Mean values of the approximate components are obtained using the DCT. Watermark is embedded according to the relationship between two adjacent mean values. Simulation and experimental results show that the algorithm has strong robustness against common audio signal processing attacks such as MP3 compression, re-quantization, re-sampling,low-pass filter and noise reduction and high efficiency.