论文标题:基于混沌的数字水印算法的研究
Study on Digital Watermarking Algorithm Based on Chaos
论文作者 高婷婷
论文导师 廖晓峰,论文学位 硕士,论文专业 计算机系统结构
论文单位 重庆大学,点击次数 90,论文页数 65页File Size3014k
2004-11-10论文网 http://www.lw23.com/lunwen_339026937/ 数字水印;混沌映射;Peano扫描;迭代混合;HVS;DCT域
digital watermark; chaotic map; Peano scanning; iterative blending;HVS; DCT domain
随着互联网和多媒体技术的普及，互联网上所传播的数字信息（例如图像、视频等）都面临着一个安全问题。这些数字产品的版权保护成了当前研究的一个热点，数字水印技术就是解决这一问题的有效方法。数字水印是在数字信息中载入一些标志版权所有者信息的标记，鉴别时通过计算机以特定的方式进行检测，以此来维护版权所有者的利益。混沌序列由于具有良好的伪随机性、轨道的不可预测性、对初值的敏感性等一系列特性，因此它非常适合于数字水印的应用。本文选择了数字图像作为水印载体，选择了由Logistic映射产生的混沌序列作为嵌入的水印信号，分别在空域和DCT域进行了水印算法的研究。本文的主要研究内容和所取得的创新成果如下：①在水印系统中采用了混沌序列作为水印信号。混沌序列对初始参数极为敏感，不同的初始值会产生不同的随机序列，可以保证水印信号的唯一性，因而攻击者伪造水印是不可能。同时，混沌序列的产生极为方便，只需给出不同的初始参数，即可产生数量众多的混沌序列，可用于需要大量水印信号的场合。③在图像迭代混合的基础上提出了空域水印算法。将混沌序列映射为一幅灰度图像，与原始图像进行迭代混合。可以证明最后得到的结果在视觉上与原图像毫无差别。仿真结果也证明了这一点。为了增强水印的鲁棒性，用Peano扫描将一个混沌序列映射为一幅图像，它保证了图像的低通特性。③一种在空域上的抗剪切的水印算法。使用一个混沌序列在原始图像中随机地选取要嵌入水印的点，由于它的随机性，可以很好地抵抗剪切攻击。结合人类视觉特性（HVS），在这些点上自适应地嵌入水印信号。算法采用了盲检测，有很强的实用性。④在DCT域，提出了一种基于分块的自适应水印算法。此算法考虑了DCT域上的AC系数的特点，并充分利用了人眼的视觉特性，在不同的分块嵌入采用不同的强度系数进行水印的嵌入，取得了较好的试验效果。本文对三种水印算法进行了大量的不可见性和鲁棒性的实验。实验结果标明，三种算法都具有良好的不可见性，对常见的图像处理操作具有较好的鲁棒性。并且方法简单易行，因此具有较高的实用价值。
With the popularization of the Internet and multimedia technology，all the digital information(such as digital image，video etc.) in the Internet are facing the security problem. How to protect the copyright of the digital products becomes one of the most attractive research issues. The digital watermarking technique is just an effective way to solve this problem. Digital watermarking can protect the benefit of the copyright owner by inserting an invisible mark into the digital information and detecting them with special methods in appropriate time. In addition, digital watermarking technique has also used to some other research fields except the copyright protecting. Chaotic sequences have good features such as noise-like, unpredictable yet deterministic orbits and sensitivity to initial condition and parameters. So chaotic signals are suitable for digital watermarking application. This paper selects digital image as watermarking objects, chaotic sequences generated by the logistic map as watermarking, and chooses spatial domain, DCT domain to investigate watermark algorithm. The main contributions in this paper are as follows:① This paper introduces chaotic sequences to the watermark systems as watermarking. The chaotic sequences are sensitive to initial parameters. Different initial parameters will generate different sequences, which can ensure digital watermarking unique. So it is almost impossible for an attacker to forge the watermark. At the meantime, the generation of chaotic sequences is very simple because different initial conditions in the attractive region will lead to different chaotic sequences, which makes it possible for chaotic sequences to be applied to the occasions with abundant watermark to be needed.② An algorithm based on iterative blending on spatial domain is proposed. A chaotic sequence is mapped to a gray image. The embedding of the watermark is done by means of the iterative blending approach. Computer simulations show that the embedded watermark is imperceptive. In order to enhance the robustness, the Peano scanning technique is used to mapping process, which maintains the low-pass property of the watermark signal.③ A crop-resistant digital watermarking algorithm on the spatial domain is also proposed. Some pixels are sought out from the original image according to a random sequence generated by a chaotic system, which makes this algorithm robust to cropping attack because of the random of the chaotic sequence. The watermark is adaptively embedded into those pixels according to the characteristic of the human visual system (HVS). Furthermore, it is also practical because the original image is not required in the detecting process.④ On the DCT domain, an adaptive algorithm based on DCT blocks is presented. The characteristic of the AC coefficients is sufficiently discussed, and based on the HVS, different inserting parameter is chosen for different block in embedding process. To shown the effectiveness of the proposed algorithm, several simulations are presented. In order to test the imperceptibility and robustness of the three algorithms mentioned above, a lot of experiments are done. Experimental results demonstrate that our schemes are effective and useful. Because of the simpleness and convenience, they may be expected to have a wide application prospect.
|
