Abstract: This paper presents a new adaptive digital audio watermark method based on discrete wavelet transform and discrete cosine transform. The watermark is embedded into the audio signal by modifying some of its coefficients in DWT domain. The selection of the coefficients is based on the time domain masking of human auditory system (HAS). Initial results obtained show that the watermarked signal is indistinguishable and satisfactory robustness against various common attacks.

Key words: Discrete wavelet transform(DWT), Discrete cosine transform(DCT), Time domain masking

摘要： 提出了一个自适应的音频水印嵌入算法。对音频载体和水印图像分别进行了离散小波变换和离散余弦变换，根据人类听觉系统的时域掩蔽效应选择嵌入点，在小波系数中的对应位置嵌入水印信号。嵌入水印后的音频信号觉察不出明显的变化，并且嵌入后的水印具有较高的鲁棒性。

关键词: 离散小波变换, 离散余弦变换, 时域掩蔽