Abstract: A lossless audio information hiding algorithm based on Tian’s algorithm is proposed.Firstly，the cover-audio is preprocessed to prevent unforeseen overflow/underflow，and then audio samples are classified into three categories.Secondly，secret information is embedded into the Least Significant Bit （LSB） of the expanded differences using the difference expansion method.After extracting the embedded message，the original audio can be perfectly restored.Experimental results show that the proposed algorithm has better hiding capacity and can efficiently improve the quality of the stego-audio.

Key words: data hiding, audio, difference expansion

摘要： 在Tian算法的基础上，提出了一种基于音频的无损信息隐藏算法。该算法首先对载体音频进行溢出预处理和采样点分类，然后利用差值扩展方法，将隐秘信息嵌入到扩展差值的LSB位。在提取时，利用无损解码技术得到位置映射表，再提取出隐秘信息后，载体音频可以实现无损恢复。仿真结果证明，算法能获得较高的隐藏容量，并能有效地保证含秘音频的感知质量。

关键词: 信息隐藏, 音频, 差值扩展