基于人工神经网络的“零数字”音频水印算法  

THE ALGORITHM OF ZERO-DIGITAL AUDIO WATERMARKING BASED ON ARTIFICIAL NEURAL NETWORKS

作　　者：郑良斌[1] 游福成[1] 何薇[1] 

机构地区：[1]北京印刷学院,北京102600

出　　处：《计算机应用与软件》2008年第6期65-67,共3页Computer Applications and Software

基　　金：北京市重点建设学科资助项目(17000024);北京市教委项目(KM200610015006)

摘　　要：提出了一种基于人工神经网络的数字音频水印算法。利用人工神经网络学习和自适应的特征,将音频信号的重要特征作为人工神经网络的输入,通过人工神经网络的学习,建立音频信号与水印信息的对应关系,达到在不修改音频信号情况下把水印"嵌入"到原始音频信号中。实验结果表明该算法具有很强的鲁棒性和抵抗常用信号处理方法处理的能力,并且在提取水印时不需要原始音频信号。A digital audio watermarking algorithm based on artificial neural networks is proposed. Utilizing the learning and self-adaptive capabilities of artificial neural networks, the relationship between audio signals and embedded watermark was learned by using the important characters of audio signal as the input of artificial neural networks, and watermark were embedded into original audio signals without modifying the data of audio. The experimental results show that the algorithm is quite robust in dealing with against normal audio signal processing, and the watermarking method does not require the use of the original audio signals for watermarking detection.

关 键 词：数字音频水印 人工神经网络 离散余弦变换 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TN912.3[自动化与计算机技术—控制科学与工程]