基于变换域的数字语音水印算法研究  

Research on Digital Audio Watermarking Algorithm Based on Transform Domain

作　　者：陈昕[1] 

机构地区：[1]福建师范大学协和学院,福建福州350007

出　　处：《宁夏大学学报（自然科学版）》2013年第4期313-316,共4页Journal of Ningxia University(Natural Science Edition)

基　　金：福建省教育厅B类科技项目(JB12270);福建省科技厅K类基金项目(JK2011007);福建省教育厅A类科技项目(JA10064)

摘　　要：结合离散小波变换和离散余弦变换在信号处理领域的优点,研究了基于变换域的数字语音水印算法.将宿主语音信号通过离散小波变换获得人耳不易察觉的低频分量后,根据均值量化的离散余弦变换直流系数正负性不易改变的特点,实现了零水印盲检测实验.结果表明,该算法有效地提高了数字语音水印的透明性、鲁棒性和抗攻击能力.Through integrating the advantages of discrete wavelet transform and discrete cosine transform in signal processing,the watermarking algorithm of digital audio based on transform domain is studied.First,the low frequency component is got by using discrete wavelet transform on the host audio signal,which the human ears are not easy to detect.Then the blind detection with zero watermarking is realized on the basis of not-easy-change plus or minus characteristic of DC coefficient with discrete cosine transform,which was first processed with mean value and quantization.The experimental results show that this algorithm can effectively improve the transparency,robustness and anti-attack ability of the digital audio watermarking.

关 键 词：离散小波变换 离散余弦变换 数字水印 语音水印 变换域 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论] TP391.42[自动化与计算机技术—计算机科学与技术]