DWT与DCT联合变换域音频水印.doc

DWT和DCT结合的变换域音频水印 摘 要 近年来随着网络通信和多媒体技术的迅速发展，盗版问题非常明显，数字水印是一种可以在开放网络环境下保护版权和认证来源的新技术。它将特定意义的标记隐藏在数字图像、音频等数字产品中，来验证创作者对其作品的所有权。 音频水印技术主要利用了人类听觉模型，在不影响音频信号质量的前提下，将水印信息隐藏在人耳不能感知的位置，来隐藏水印数据。本文研究DWT和DCT结合的音频水印算法，首先对音频信号进行DWT和DCT变换，根据水印的大小把DCT系数均匀分段，采用量化的方法把直观的二值黑白图像嵌入音频中，然后利用攻击函数攻击音频信号，再从含水印的音频中提取出水印信息，最后通过MATLAB仿真计算信噪比和归一化系数，验证该算法是否具有良好的安全性、鲁棒性和不可感知性。 关键字 音频水印，离散小波变换，离散余弦变换 ABSTRACT In recent years, with the development of the network communication and multimedia technology, the problem of piracy is becoming more and more obvious. Digital watermarking is a new kind of technology, which can be used to protect copyright and certificate source under the opening network environment. The technology will hide the special markers in the digital image、audio to prove the creators ownership of their works. On the condition of not affecting the quality of audio signal, audio watermarking technology mainly uses the human hearing model to hide the watermarking information in the position where peoples eyes can not perceive. The paper studied the audio watermarking algorithm based on DWT and DCT. Firstly made DWT and DCT transformation, then DCT coefficients were segmented equally according to the size of watermark. By using the quantification method, intuitive binary black and white images were embeded into the audio signal. Later attack functions are used to attack the audio signal and then extract the watermarking information. At last though the MATLAB simulation, calculate signal to noise ratio and normalized coefficient to prove whether the algorithm satisfy the requirements of imperceptibility, safety and robustness or not. Key word: Audio watermark, Discrete wavelet transform, Discrete cosine transform  目录 TOC \o 1-3 \h \u HYPERLINK \l _Toc14346 1. 绪论 PAGEREF _Toc14346 1 HYPERLINK \l _Toc14469 1.1 音频水印的发展 PAGEREF _Toc14469 1 HYPERLINK \l _Toc30734 1.2 数字水印原理及简单模型 PAGEREF _Toc30734 1 HYPERLINK \l _Toc20154 1.3 数字音频水印存在的问题 PAGEREF _Toc20154 3 HYPERLINK \l _Toc1