基于语音低频分量的语音零水印  

Audio Zero Watermark Based on The Multi-level Vector Quantization of Linear predictive coding coefficient

作　　者：郑锦波[1] 张贵仓[1] ZHENG Jin-bo, ZHANG Gui-cang （College of Mathematics and Informational Science, Northwest Normal University, Lanzhou 730070, China）

机构地区：[1]西北师范大学数学与信息科学学院,甘肃兰州730070

出　　处：《电脑知识与技术》2010年第12期9849-9851,共3页Computer Knowledge and Technology

摘　　要：提出一种基于语音低频分量的数字零水印算法。通过对语音信号进行二级小波分解，之后提取最前面的小波细节分量并利用小波细节分量的混合域系数和水印运算得到检测密钥，并在接收端进行盲水印提取．实验结果证明，该算法对于各种攻击尤其对幅度缩放、重量化和低通滤波具有较好的鲁棒性和不可见性。An audio zero watermark based on the Multi-level Vector Quantization of Linear predictive coding coeffcient method was proposed. In the proposed method, we can get the key with the method that the robust watermark were embedded in The Multi level Vector Quantization of Linear predictive coding coefficient which is pick up from the audio and index constrained respectively. Simulation results demonstrate the effectiveness of the proposed algorithm in terms of robustness and imperceptibility in amplitude scaling, low-pass filter and volume normalization.

关 键 词：数字零水印 语音低频分量 鲁棒性 版权确认 

分 类 号：TP309[自动化与计算机技术—计算机系统结构;自动化与计算机技术—计算机科学与技术]