一种基于基音周期估计的音频水印算法  

An audio watermark algorithm based on pitch estimation

作　　者：付永庆[1] 孙滢[1] 

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《应用科技》2007年第12期4-7,共4页Applied Science and Technology

摘　　要：提出了一种水印信号与载体信号均为语音的数字音频水印算法.主要由用混沌序列对音频水印加密,基于载体信号基音周期估计确定水印嵌入位置和根据量化步长Δ确定水印嵌入强度等关键步骤所组成.仿真结果表明,该算法不仅具有较好的透明性,而且也对叠加噪声、低通滤波、重新采样等攻击具有较好的鲁棒性.A digital audio watermarking algorithm is presented in which both watermarking signal and original signal are audio. In the method, the audio watermarking is encrypted using chaotic sequences, the embedding position is determined in accordance with pitch estimation of the speech signals, and the embedded watermarking intensity is determined according to the quantization step size. Simulation results show that the proposed watermarking scheme possesses perfect transparency and good robustness against general signal-processing operations such as noise addi- tion, low pass filtering, and re-sampling.

关 键 词：音频水印 混沌加密 基音周期估计 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]