数字音频双重水印算法  

A Digital Audio Dual-Watermarking Algorithm

作　　者：陈金儿[1] 王让定[1] 王晓丽[1] 

机构地区：[1]宁波大学信息科学与工程学院,浙江宁波315211

出　　处：《宁波大学学报（理工版）》2006年第1期54-58,共5页Journal of Ningbo University：Natural Science and Engineering Edition

基　　金：浙江省自然科学基金(Y104144);宁波市博士基金(2005A610003)

摘　　要：提出了一种同时嵌入鲁棒水印和易损水印的算法.利用离散小波变换的多分辨率特性以及离散余弦变换的解相关和能量压缩能力,在中低频区嵌入鲁棒水印;在小波域的高频区嵌入易损水印.水印的嵌入采用量化的方法,提取水印信号时不需要原始音频信号.实验结果表明,文中算法的水印鲁棒性高,适于音频作品的版权保护;而文中的易损水印对音频攻击具有很强的敏感性,可同时认证音频作品的内容完整性.In this research an audio dual -watermarking scheme is presented in which both robust watermark and fragile watermark can be embedded. The robust watermarking is embedded in low-middle frequency range using the multi-resolution of DWT and the energy compression of DCT, while the fragile watermarking is rooted in high frequency range of wavelet domain. Watermark embedding is based on quantization process. The watermark can be extracted without the original digital audio signal. Experimental results show that the robustness of robust watermark is satisfying, and it can be applied to copyright protection, while fragile watermark is sensitive to audio attacks and can thus be utilized for content integrity verification.

关 键 词：数字音频 双重水印 盲检 

分 类 号：TN912.3[电子电信—通信与信息系统] TP309.2[电子电信—信息与通信工程]