Digital Watermarking Embedding Technology
Based on Fourier Transform
ABSTRACT
Along with the computer and network technology rapid development, and dissemination of digital works have become increasingly easy to copy, simultaneously allow digital works to protect the security of information and copyright protection has become an urgent need to address the real issues. In recent years, digital watermarking is in the field of information security emerging intellectual property protection method. It passed in the original data embedded in some important information protected by copyright for the data integrity of media ownership and attribution mention for complete and reliable evidence, thereby to prevent the piracy of digital products and tampering with purpose.
This paper presents a Fourier-domain based on the meaningful watermarking algorithm, mainly in the following:
(1) To enhance the security watermark, before the watermark embedded watermark on the first use of
(2) To improve the traditional detection methods related to the accuracy, used embedded watermark embedding two are not related to the pseudo-random sequence, greatly improve the detection accuracy.
(3) low-frequency part of the image most concentrated energy; changes here easily lead to distortion, therefore this paper, revising part of the high-frequency method for embedding.
(4) This watermark detection method for the semi-blind testing, only part of the original image information in, therefore have important practical significance.
Key Words：Digital watermarking；Fourier transform；Image processing；
Information security
基于傅立叶变换的数字水印嵌入技术
摘 要
随着计算机及网络技术的飞速发展，数字作品传播和拷贝变得越来越方便，同时使得数字作品的信息安全保护和版权保护也成为迫切需要解决的实际问题。数字水印是近年来在信息安全领域兴起的保护知识产权的新方法。它通过在原始数据中嵌入一些重要信息为受到版权保护的媒体数据的完整性和所有权归属提供完全和可靠的证据，以此达到防止数字产品的盗版和篡改目的。
本文提出了一种基于傅立叶域的有意义水印算法,主要做了如下工作：
(1)为了提高水印的安全性，在嵌入水印之前首先对水印信息利用Arnold变换进行置乱。
(2)为了提高传统相关性检测方法的准确率，嵌入水印时采用嵌入两个不相关伪随机序列的方法，大大提高了检测的准确率。
(3)低频部分集中图像的大部分能量，修改这里容易引起失真，所以本文采用修改中高频部分的方法进行嵌入。
(4)本文水印的检测方法为半盲检测，只需要原始水印的部分信息的参与，所以具有重要的现实意义。
关键字：数字水印；傅立叶变换；图像处理；信息安全1173
[1] [2] [3] [4] [5] [6] [7] [8] 下一页