基于DWT和相对较小值的音频盲水印新算法  

Effective and Relatively Small Values of DWT-Based Audio Blind Watermarking Algorithm

作　　者：朱有忠[1] 向俊[2] 

机构地区：[1]吉林省统计局数据管理中心,长春130051 [2]吉林大学软件学院,长春130012

出　　处：《吉林大学学报（信息科学版）》2010年第2期203-208,共6页Journal of Jilin University（Information Science Edition）

摘　　要：为了解决网络环境下的音频数字产品的版权保护问题,提出了一种新的基于DWT(Discrete Wavelet Transform)和系数值比较的音频盲水印新算法。该算法通过相邻3个水印比特组合的二进制值之和确定其嵌入位置,并把其系数修改为相邻系数中较小的值,嵌入位置选取在低频系数上,因此具有很强的抵御各种攻击和自适应能力,适用于数字音频等信号的版权保护。算法实现快速、简单有效。DWT的使用提高了水印的不可感知性,通过Arnold置乱不但保证了水印图像的安全性,而且提高了水印的鲁棒性。在Matlab上的仿真结果表明,该算法具有较强的鲁棒性和透明性。In order to solve the problem of audio digital products copyright protection under the network environment, a new algorithm for digital audio watermarking based on the DWT （Discrete Wavelet Transform） is put forward. The algorithm uses the sum Binary value of the neighboring three watermark bit combination to locate the embedding position, and make its value smaller. The watermarking picture being embedded into audio signal also ensures the ability of attack resisting and adaptation, which could be used to protect the copyright of digital audio signals. The merit of the algorithm is that it takes advantage of the speediness and simplicity of the DWT. Arnold transformation guarantees the security of watermark image. The simulation experiment on the MATLAB is carried based on the algorithm and the results demonstrate robustness and imperceptivity of the algorithm.

关 键 词：离散小波变换 置乱 数字音频水印 相对较小值 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]