【摘要】 提出一种小波域自适应量化的语音伪装算法。首先对语音信号进行小波变换,由混沌序列控制嵌入位置,并采用固定阶距量化实现秘密信息的隐藏。在此基础上通过F＿范数自适应控制量化阶距的变化,实现了基于F＿范数的语音伪装算法。仿真实验表明算法安全性高、透明性好,在多种干扰情况下均能盲提取出秘密信息。更多还原

【Abstract】 This paper presents a speech camouflage algorithm based on adaptive quantization in wavelet domain. First speech is segmented to frames and performed discrete wavelet transform. Then secret information is hidden by constant step quantization and the quantized wavelet coefficients are determined by the chaotic sequence. Furthermore, the speech camouflage algorithm is performed based on frobenius norm which adaptively controls the quantization steps. Simulations show that the algorithm is secure, imperceptible. Despite of distortions, the secret information can be extracted without the original speech.更多还原