一种小波域音频信息隐藏方法

提出了一种基于量化的小波域音频隐藏算法，将保密语音隐藏到载体音频中．为提高隐藏重和保密语音传输的安全性，对保密语音进行了小波域压缩编码和m序列的扩频调制，生成待隐藏的比特序列；通过量化方法，将编码和调制后的保密语音隐藏到载体音频的小波系数中；保密语音的恢复过程不需要使用原始音频、仿真结果表明，隐藏有保密语音的载体音频听觉质量没有明显下降，提取的保密语音感知质量较好；该算法对重量化、加噪、低通滤波等攻击均有良好的鲁棒性．

An Audio Information Hiding Technique in Wavelet Domain

An audio information hiding algorithm in wavelet domain based on quantization is presented. The secret speech is hidden in the carrier audio. The coded bit sequences, which are obtained from the compressed speech, are modulated with the m-sequences. These modulated sequences are embedded into the DWT coefficients of the carrier audio via quantization. The original carrier origin is not required in recovery processing. Experimental results show that the quality degradation of the carrier audio is unperceivable, and the quality of the recovered speech is satisfying. This algorithm is robust to many attacks, such as requantization, noise-addition, low-pass filtering and so on.