Abstract: A novel data hiding method for digital audio using integer wavelet transform and the hearing threshold embedding technique is proposed.Data are embedded into the Least Significant Bit-plane（LSB） of integer wavelet coefficients，which accords with the principle of Human Auditory System（HAS）.The inverse integer wavelet transform is utilized to modified coefficients to form a new audio sequence（stegosignal）.Experimental results show that this method is large payload，high audio quality and full recovery.The method is successfully applied in the six types of digital audios.

Key words: audio steganography, adaptive data embedding, integer wavelet transform, auditory threshold, Least Significant Bit-plane（LSB）

摘要： 提出了一种新型数字音频信息加密算法。通过整数小波变换，将加密数据嵌入到载体音频的小波子带系数中，在整数域估算听觉阈值，并将之作为嵌入信息的门限值，最后采用整数小波变换的逆变换来修正小波系数从而形成一个新的音频序列信号。该算法具有较大的有效载荷，较好的音频质量，可以实现音频信息的完全重构。通过对6种格式的音频信号进行测试，证明了提出的方法是有效的。

关键词: 数字音频加密, 自适应数据嵌入, 整数小波变换, 听觉阈值, 最低比特位