基于DWT的双重数字音频盲水印算法  

An Algorithm of Dual Digital Audio Blind Watermarking Based on DWT

作　　者：任克强[1] 李辉环[2] 谢斌[1] 

机构地区：[1]江西理工大学信息工程学院,江西赣州341000 [2]南昌大学共青学院信息工程系,江西九江332020

出　　处：《江西理工大学学报》2011年第3期1-5,共5页Journal of Jiangxi University of Science and Technology

摘　　要：针对单一水印存在的问题,设计了一种基于DWT的双重数字音频盲水印算法,先用混沌序列对水印图像加密,再将原始音频信号分段,采用不同的嵌入策略平衡水印透明性和鲁棒性之间的矛盾,把水印信息分别嵌入到经过四层小波变换的高、中频分量中,水印的提取不需要原始音频信号和原始水印,实现了水印的盲检测.实验结果表明,水印的透明性很好,在抵抗重采样、滤波、噪声干扰以及压缩等多种攻击时具有良好的鲁棒性.Aiming at the existing problem of single watermarking,this article has designed an algorithm of dual digital audio blind watermarking based on DWT.The watermarking is encrypted by the chaotic sequence,then the original audio signal is divided into segments,the different watermarking embedding strategy is used to balance the contradiction between watermarking transparency and watermarking robustness,the watermarking information is respectively embedded into the high frequency component and the medium frequency component through four-level wavelet transform.The watermarking extraction does not need the original audio signal and the original watermarking,the watermarking blind detection is realized.Experimental results show that the watermarking transparency is very good,and it also has good robustness against many kinds of attacks such as resampling,filtering,noise interference,compression and so on.

关 键 词：音频水印 双重水印 盲水印 混沌加密 离散小波变换 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]