摘要
针对数字音频水印盲提取中存在的算法复杂、稳健性差等问题，提出一种基于快速独立分量分析的音频水印方案。研究中分别采用二维图像及特定音频信号作为水印信息，通过嵌入位置随机选取、水印预处理、载体音频DCT变换等方法，提高水印的透明性及鲁棒性。仿真实验显示，该方案算法简单、效率高，在不同攻击情况下均具有较好的稳健性。
To solve the problems o{ the complexity and robustness of some blind digital audio watermark technologies, an algorithm based on Fast ICA is presented. The picture and audio signals are used as the watermark to embed in the original audio. The watermark is robust and imperceptible by selecting the embedded point randomly and transforming the original audio by DCT. The experimental results show that the algorithm has good robustness against common attacks.
出处
《武警工程大学学报》
2015年第6期44-47,共4页
Journal of Engineering University of the Chinese People＇s Armed Police Force
