摘要
阐述了音频水印技术的嵌入及提取过程,提出了一种将二值图像嵌入音频信号的数字水印算法.该算法可对二维数字水印图像置乱处理并降维形成一维水印序列;对数字音频信号分帧并做倒谱分析,利用倒谱域两端因数比较大,其余的幅值都在0附近波动的特点,采用统计均值调制的方法,在倒谱域嵌入水印.实验结果表明,该方法具有良好的不可感知性和鲁棒性.
In this paper, we propose a novel robust audio watermarking scheme in spectrum. Because most of audio spectrum coefficient value are around zero except for two sides, we propose to embed data by manipulating statistical mean of selected spectrum coefficients. Our experiment results have shown that the novel audio watermarking scheme in spectrum domain can achieve imperceptibleness and robustness
出处
《大庆石油学院学报》
CAS
北大核心
2005年第4期130-132,共3页
Journal of Daqing Petroleum Institute
关键词
音频水印
倒谱分析
统计均值调制
audio watermarking
reversed spectrum analysis
statistical mean manipulation
