【作者】 朱桂斌；

【导师】 曹长修；

【作者基本信息】 重庆大学 ， 控制理论与控制工程， 2004， 博士

【摘要】 在现在大多数的通信系统中，人们对通信内容的保密往往求助于密码学。然而，在许多领域，密码学的应用已经越来越明显地暴露出它的局限性：密码学通过密文的不可理解性来保护信息的内容，而密文的不可理解性同时也暴露了信息的重要性。这很容易引起攻击者的注意，从而吸引攻击者采取多种手段对通信的内容进行破译或对通信过程进行破坏，继而造成通信过程的失败。近年来，一种新的信息保密方式成为信息安全领域的一个热门研究课题，这就是信息隐藏。与密码学不同的是，它通过对信息存在的本身或信息存在的位置进行保密，从而提供了比密码学更加可靠的安全性。 对于信息隐藏技术的研究最为关心的三个主要方面是：抗检测性(不可感知性)，隐藏容量和鲁棒性，这三者之间是互相冲突的，不可能同时达到最优。在具体应用时必须在三者之间寻求某种平衡，对于一般信息的隐藏而言(如图像信息)，应努力寻求高的不可感知性和隐藏容量，而牺牲其鲁棒性。对于秘密信息的隐藏而言，应努力寻求高的不可感知性(包括不可检测性)和鲁棒性，而牺牲其隐藏容量。本论文在信息隐藏方面的主要工作就是围绕这两种应用展开的，内容涉及图像置乱、图像信息隐藏和隐藏分析和攻击技术等方面。 本论文在理论上所作的主要工作及所取得的结论如下： ①系统地研究了整数的处理技术 在第4章对可逆的线性变换进行了改造，使之成为整数到整数的变换。首先介绍了三类基本的整数可逆变换，在此基础之上，通过对给定线性变换的矩阵分解，给出了一个可逆线性变换整型化的充要条件及其构造方法，并将变换的应用范围从实数域推广到复数域。变换本身是可逆的，因此非常适合于无失真的数据处理，如语音或图像的无损压缩和加密处理。 ②研究了仿射模变换的构造问题 在第3章提出了广义Fibonacci序列的概念，在此基础上，成功地解决了整系数的仿射模变换的构造问题。 本论文在应用上所作的主要工作及所取得结论如下： ①关于图像置乱算法 简要概述了图像置乱的研究现状及研究意义，给出了图像置乱、置乱周期和图像置乱程度的数学定义。根据对待图像的不同观点，提出了3种新的图像置乱算法：准逆序变换的置乱算法，基于准抖动变换的置乱算法，基于拟仿射变换的置乱算法，并分析了它们在图像置乱中的性能。针对前人提出的仿射模变换的置乱算法，成功地解决了它的构造问题。重庆大学博士学位论文 基于准逆序变换和准抖动变换的置乱算法的优点是:如结合颜色空间将灰度值同时置乱，则置乱效果与均匀分布的随噪声一样，用于隐藏可增强隐蔽通信的抗统计分析的能力。 基于拟仿射变换置乱算法的优点是:可控参数是5个随机的浮点数，通过变换的很少次数的迭代可达到理想置乱效果。 ②数字图像中信息隐藏的算法 简要介绍了信息隐藏技术的分类及基本算法，着重研究了两类以数字图像为载体的典型信息隐藏系统:图像信息的隐藏算法和一般数据信息的隐藏算法。前者要求隐藏算法有较大的隐藏容量，后者对隐藏算法的鲁棒性方面有更高的要求。 对于图像信息的隐藏系统，除要求有较大的隐藏容量外，“盲提取”是一项十分重要的要求。“盲提取”可以使通信过程一次完成，不必依赖其它的通信过程，大大简化通信过程。为了实现隐藏信息的“盲提取，，’在研究基本系统的。‘盲提取”性能的基础上，对基本系统进行了改进，大大提高了系统的隐藏信息的“盲提取”性能。 对于一般数字信息的隐藏系统，要求隐藏的数据完全能够提取出来，要有比较强的鲁棒性。由于95%以上的图像信息都是以压缩方式进行存储和传输的，所以将数据压缩和信息隐藏结合在一起进行研究，提出了基于数据压缩思想的通用信息隐藏模型。针对模型中的关键技术进行研究，解决了整型图像数据的可逆浮点运算问题，实现了在不增加数据量和不减少信息量的条件下，改变数据的统计特性，增强了系统的抗检测性。 ③信息隐藏分析及攻击技术 对目前国际国内这方面的研究进行了概述和总结，简要介绍了密码分析技术和和隐藏检测技术，介绍了信息隐藏的检测与攻击一般框架以及数字图像信息隐藏检测技术和数字水印攻击方法。关键词:信息隐藏，图像置乱，图像加密，隐蔽通信，隐藏分析，数字水印更多还原

【Abstract】 In the most of modern communication systems, the security of the traffic primarily depends on the cryptography, but this has rarely been adequate in practice by merely concealing the content of a message using encryption because the encryption may attract the attacker’s attention, and then invokes attack. Information hiding techniques have recently become important in a number of application areas. In contrast to cryptography, it seeks to conceal the sender, the receiver or the very existence of a message, rather than merely concealing its content.The most concerned aspects of information hiding are, as to say, the ability to endure detecting or imperceptibility, the embedding capacity and the robustness, In general, these three aspects of information hiding are inconsistent to each other and can’t get perfect performance independently. In practice, the imperceptibility and capacity are more important when you want to embed average information such as a picture, while you want to hide important information such as military applications, the robustness is more important. In this dissertation, we study the technologies and application of these two situations, our attentions focused on image scramble, image hiding, hiding analysis and attack.The contributions of this dissertation in theory include:(1) The integer version of data processing is studied systematically.A method to construct an integer version of the linear transforms is presented so that the integers of input data can be mapped to integers. Based on the introduction of three kinds of basic integer reversible transform, it is proposed by means of matrix decomposition the sufficient and essential condition that a linear transform can be rebuilt into integer version and the corresponding construction method. The transform is reversible, so it is suitable for the lossless data processing such as the lossless digital audio or image compressions.(2) The construction of module affine transformation.The concept of Generalized Fibonacci Queue is presented, based on which, the construction of module affine transformation is resolved.The contributions of this thesis in application include:(1) Digital image scrambling techniquesAs a kind of encryption method, the importance of image scrambling technique in information hiding is explored. The definitions of image scrambling, image scramblingdegree and periods of image scrambling are proposed. Three novel image-scrambling algorithms are proposed, and their performances of scrambling image are analyzed.The quasi-anti-order scrambling and quasi-dither scrambling have ideal scramble performance, and can enhance the anti-analysis performance of an information hiding system.Quasi-Affine Transformation Over Limited Integer Grids (QATLIG) is present, based on the idea of building reversible linear transform that mapping integers to integers. QATLIG has a larger periods and leads to a better result, so it is superior to Arnold transformation, Fibonacci transformation and other geometric transformations.(2) Information hiding technique in digital imageThe classification and the basic algorithms of information hiding are introduced. We focused on two kinds of applications, hiding algorithms for image embedding and algorithms for general important information embedding. In general, the first one request high-embedding capacity but the second demands robustness.In practice, the ’blind extraction’ is another demand for image embedding, which ensure the simplicity of the communication procedure. We propose a new algorithm, which enhance the blind performance greatly.As above 95% of images are compressed for storage and transfer, data compression techniques are combined with information hiding algorithms. A general information-hiding model based on data compression is proposed, we researched the key problems in this model, change the statistic properties of embedded data under the restriction of no data inflation and enhance the ability of anti-detection.更多还原