摘要
提出了一种基于DWT变换域的数字音频水印算法，对小波变换系数采取模2取余方法作为数字水印嵌入策略。通过仿真实验结果表明主要具有以下几个方面的特点：(1)水印隐蔽效果好，凭借人的感觉系统无法分辨出与原始音频的差别；(2)可以有效地抵抗噪声、低通滤波、有损压缩的处理；(3)在检测水印要受到密钥的限制，不知道口令的人是无法正确恢复出水印图像的。另外本算法在检测水印时不需要原始的音频信号。
The paper proposes a watermarking algorithm to embed copyright protection into digital audio based on modular arithmetic in DWT. Extensive experimental results illustrate that the technique significantly possesses robustness to against signal processing. Main features of the proposed method are: (1) The difference of watermarked audio and original is imperceptible. (2) The watermark embedded into audio is an image. It is more near the application of the fact compared with the random sequence. (3)The embedded watermark can be extracted only by a secret key. In addition, Extraction of embedded watermark does not need original audio.
出处
《计算机工程》
CAS
CSCD
北大核心
2004年第6期44-45,52,共3页
Computer Engineering
基金
国家重点基础研究发展计划(973)基金项目(G1998030509)
