节点文献

信息隐藏和隐藏分析的理论与算法研究

Theory and Algorithm Research about Information Hiding and Steganalysis

【作者】 姜楠

【导师】 胡正名; 杨义先;

【作者基本信息】 北京邮电大学 , 密码学, 2006, 博士

【摘要】 随着数字技术和网络技术的不断发展,人们可以方便、迅速地将数字信息传达到世界各地。如何解决与此相关的版权保护、完整性认证、秘密通信等问题,成为亟待解决的研究课题。信息隐藏就是为解决这些问题而出现的,它有两个主要分支:数字水印和隐写术。数字水印主要用于版权保护、电子文档完整性保护等方面。隐写术主要用于秘密通信。同时为了保卫国家安全,信息隐藏分析也逐渐发展起来。 本文在国家重点基础研究发展规划973项目(TG1999035804)、国家自然科学基金项目(60473016)和北京邮电大学研究生创新基金项目的资助下,围绕与信息隐藏和信息隐藏分析相关的算法和理论问题,进行了研究。主要工作如下: 1.针对不同的应用环境,设计了3种不同的数字水印算法。自适应HSI鲁棒彩色图像数字水印算法将图像变换到HSI色彩模型中,并利用人眼视觉特点计算临界差异,在不可见性和鲁棒性之间找到折衷点。强鲁棒图像数字水印算法利用傅里叶变换的几何不变域,可以抵抗长宽不等比缩放、基本旋转等RST攻击,还可以抵抗格式转换、常规信号处理等攻击。同时能够实现嵌入量大、盲检测、100%正确提取等要求。Midi音频数字水印算法利用Midi文件格式的特殊性,借鉴软件水印的方法,实现了水印算法完全的透明性。 2.提出自适应加性水印嵌入模型,在此前提下,用PSNR衡量不可见性、用信道误码率衡量鲁棒性,给出了不可见性、鲁棒性和信息隐藏率之间的定量关系,指出达到三者折衷点的关键是水印嵌入强度。实验指出如何在确定水印算法参数之后计算嵌入强度,以达到三者的折衷。 3.指出纠错码在水印信道中的作用。水印信道受不可见性影响,嵌入强度与水印长度的乘积是一个常数(此常数的值与水印算法要达到的不可见性、鲁棒性等有关),这是水印信道与一般通信信道最大的不同。因此水印信道中纠错码的码率不能任意小,文中给出了纠错码码率的下限。 4.在分析已有信息隐藏模型不足的基础上,给出了新的信息隐藏模

【Abstract】 With the development of digital technology and networking technology, people would quickly and easily communicate with each other through the network. Solving the correlative problems, such as copyright protection, intergrality authentication, privacy communication and etc., is becoming more and more urgent. Information hiding can solve these problems. It has two branches: digital watermarking and steganography. Digital watermarking is used main for copyright protection and intergrality authentication. Steganography is used main for privacy communication. At the same time, in order to protect the national security, steganalysis is developed gradually.This dissertation systemically researches the theory and the algorithm about information hiding and steganalysis. It is sponsored by the National Grand Fundamental Research 973 Program of China (No. TG1999035804), the National Natural Science Foundation of China (No. 60473016) and the Graduate Students Innovation Foundation of BUPT. The main work is as following:1. Designs three different digital watermarking algorithms for different application environments. The first algorithm is named self-adaptive HSI robust color images algorithm. It transforms images into HSI color model and gains the just noticed difference using the characteristic of human visual system. The second algorithm is called high robust images algorithm. Using the geometric invariable region, it can resist zooming in two rates of length and width, basic circumvolving, filtering, adding noise, format changing and etc. The merits of the algorithm also include: a) the hiding rate is very high; b) blind extraction; c) bits error rate is zero. The third algorithm is called

节点文献中: 

本文链接的文献网络图示:

本文的引文网络