基于混沌序列的小波域语音信息隐藏方法  

Speech Information Hiding Approach in Wavelet Domain Based on Chaotic Sequence

作　　者：张晓明[1] 殷雄[1] 

机构地区：[1]北京石油化工学院信息工程学院,北京102617

出　　处：《系统仿真学报》2007年第9期2113-2117,共5页Journal of System Simulation

基　　金：北京市教委科技发展计划面上项目(KM200510017006)

摘　　要：提出一种将保密语音隐藏到音频载体的离散小波变换新方法,并引入Logistic混沌映射,利用混沌序列对秘密语音信息进行处理。提出了详细的隐藏算法描述,对音频载体进行4级小波变换并分段,然后将秘密语音嵌入到小波变换系数的中高频部分。在算法中引入了嵌入强度的概念和小波系数的能量分配比值,且使秘密信息具有盲提取效果。实验结果表明,选择合理的嵌入强度值,本算法在不同类型的载体下都能够保证优异的隐藏效果。该算法在受到重采样和回声干扰等攻击时具有很强的鲁棒性。A kind of new approach of hiding speech into the audio carrier was proposed in discrete wavelet transform （DWT） domain. Logistic chaotic map was utilized as chaotic sequence to process the secure speech. The hiding algorithm was described as embedding and extracting processes. The audio carrier was performed in four-level DWT and converted into segments. The secure speech was embedded into the medium and high frequencies of the wavelet efficiency. An embedding strength was constructed in the algorithm, along with an energy ratio of wavelet efficiency. The secret information can be extracted blindly. Experimental results show that under the different several audio carriers with suitable embedding strength value, the hiding effect is very satisfying. Besides, the algorithm is robust to attacks of re-sampling and echo interfacing.

关 键 词：语音 信息隐藏 离散小波变换 混沌序列 鲁棒性 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]