一种基于人耳掩蔽效应的DCT域音频水印

A DCT Domain Audio Watermarking Algorithm Based on Human Auditory Masking Properties

文中提出了一种基于听觉掩蔽效应的DCT域音频水印算法。首先根据人耳听觉系统的心理声学模型估算每帧音频数据的掩蔽阈值,然后利用量化的思想将水印信号自适应地嵌入到掩蔽阈值以下的频率分量所对应的DCT域系数中。仿真实验结果表明:与传统DFT方法相比,该算法具有更好的透明性和鲁棒性。

This paper presents an audio watermarking algorithm in DCT domain based on human auditory masking properties. First the masking threshold of each frame of audio signal is calculated according to psychoacoustic model, and then the watermark signal is adaptively embedded in the DCT coefficients whose corresponding frequency components are under the masking thresholds by using the quantification method. Simulation results indicate that： as compared with the traditional DFT method, the algorithm has better transparency and robustness.