摘要： 提出一种基于小波去噪技术和小波突变点检测技术的音频扩频掩密分析算法。该算法利用小波去噪算法分离含密音频上的噪声信号，循环截取不同长度的噪声信号与剩余的噪声信号进行滑动相关计算，利用小波突变点检测技术检测计算得出的滑动相关值中的突变点，提取这些突变点特征对待检测的音频进行掩密分析。实验结果表明，在PN序列嵌入强度大于0.002时，算法的检测正确率达80%以上。

关键词: 音频掩密分析, 扩频, 掩密, 小波变换

Abstract: This paper proposes a new audio spread spectrum steganalysis algorithm based on wavelet denoising technology and wavelet discontinuity detection technology. The algorithm uses the wavelet denoising technology to separate the noise signal from the audio, and circularly interceptes a part of the noise signal with different length. Using this part of noise signal, it computes the cross correlation sequence with the rest of the noise signal. Using wavelet discontinuity detection technology, it detects the discontinuities from the cross correlation sequence, extracts some of the characteristics from those discontinuities to make steganalysis of audio to be detected. Experimental results show that the correct detection rate of the scheme achieves above 80 percent when the strength of the pseudorandom noise sequence is above 0.002.

Key words: audio steganalysis, spread spectrum, steganography, wavelet transform

中图分类号: 

• TP391