摘要
提出了一种在话音传递中,蒙蔽人类听觉的信息隐藏模型。给出了数字音频信号中信息隐藏的伪随机序列生成算法。利用此算法对隐蔽信息先加密,再对数字音频信号进行分段处理,并选出部分段做离散余弦变换(DCT),最后依据人类的听觉系统选择DCT域中高频系数进行量化完成加密的隐蔽信息嵌入。实验结果表明,在提高鲁棒性的同时,还大大增加了安全性,就是在外界发现数字音频含有隐藏信息的情况下,在不知道密钥的情况下,还是很难提取隐蔽信息。
An information hiding model is introduced which can becloud human auditory in transmission of tone. The information is hidden in digital audio signals by an arithmetic of Preudo-random sequences generation. Encrypted the hidden information, then processed the digital audio signal in subsections and carryed out discrete cosine transform (DCT), with some parts of section. Selected the high frequency coefficient in DCT domain according to human auditory system to be quantified as to embed the encrypted information hiding. As a result of experiment, both security and robustness are improved greatly at the same time. Even if others discovere there exists information hiding in the digital audio, its difficult to abstract the hiding information without the cryptographic key.
出处
《计算机工程》
CAS
CSCD
北大核心
2004年第22期19-20,26,共3页
Computer Engineering
基金
国家自然科学基金项目(60271025
60271022)
