基于迭代相位恢复算法的光学多幅图像认证系统的研究与设计

发布时间：2018-04-30 21:40

本文选题：光学信息安全 + 信息光学　；参考：《山东大学》2017年硕士论文

【摘要】：随着网络技术,计算机通信技术、密码科学和多媒体通信技术的飞速发展,现代社会已步入信息大量传播和飞速发展时代,信息的安全传输、保管和版权问题也来越受到当代社会的重视。自从傅里叶域的双随机相位加密技术用来对数字图像进行加密和隐藏以后,基于信息光学的多媒体数字信息加密方法已变成一个全新的研究方向,并且吸引了越来越多的国内外研究机构和研究员投入这一领域中。基于信息光学的图像加解密系统通常利用信息光学技术诸如Fourier transform、fractional Fourier transform 等技术和相关技术诸如 compressive sensing、computational ghost imaging等方法,对待加密信息进行加密解密、数字水印的隐藏和认证。与传统的安全系统相对比,基于光学和光子学方法的安全系统的优点在于光波具有更复杂的自由度例如强度、相位、偏振、非线性变换和光子的量子性质等,此外光学信息可以进行复用,使得安全系统更加安全和难以破解。本论文将首先对国内外最新的加解密和水印技术进行简单的介绍,系统梳理常见的光学图像加密技术。在此基础上,本文将借助迭代相位恢复算法并结合复用技术和稀疏数据表示等技术展开研究工作,并提出新颖、可行的改进方案。具体研究内容包括以下几个方面:(1)将迭代相位恢复算法与纯相位掩膜复用技术结合,对多幅图像进行加密,实现了用同一套加密图像对多幅图像的同时解密,可以极大地提高加密图像的利用率。(2)结合迭代相位恢复算法、空间位置复用和稀疏数据表示等技术,对经过迭代相位恢复算法计算得到的变换平面的相位部分进行空间位置复用,并进行了数据模拟仿真。试验结果表明通过空间位置复用技术得到的解密图像虽不能直接通过肉眼或相关系数来进行分辨,但能通过非线性相关系数进行分辨,在显著地降低了处理和传输的数据量的同时,大大地提高了系统的安全性。(3)提出一种基于信息光学理论的级联分级多幅图像编码和认证方案。本论文拟采用迭代相位恢复算法、纯相位掩模复用、空间位置复用以及稀疏数据表示等技术,设计实现一种光学分级多幅图像编码和认证方案。首先通过逆菲涅耳衍射将第一级认证图像编码为复振幅场,然后对其振幅部分取样并求和得到复合图像的振幅部分;以得到的复合图像的振幅为输出振幅限制,通过相位恢复算法将第二级认证图像编码为纯相位掩模,通过纯相位掩模复用技术可以得到最后的复合图像;通过迭代相位恢复算法对复合图像进一步编码得到最后的纯相位密文;最后通过特定的公式分别设计两级图像认证时的解密密钥。认证过程中只需要在相应的位置放入相应的解密密钥,通过非线性相关分布和相关系数两个不同的判据,可以完成对多幅图像的不同级次的认证和识别。本方案在常见的光学多幅图像编码及认证的基础上,通过稀疏表示和相位复用等技术不仅增加了编码图像的数量,而且实现了分级认证。
[Abstract]:With the rapid development of network technology, computer communication technology, cipher science and multimedia communication technology, modern society has entered a era of information dissemination and rapid development. The security transmission of information, storage and copyright issues have also been paid more attention by the contemporary society. The dual random phase encryption technology has been used in the Fourier domain. After the image is encrypted and hidden, the multimedia digital information encryption method based on information optics has become a new research direction, and attracts more and more domestic and foreign research institutions and researchers to invest in this field. Information optics based image encryption and decryption systems usually use information optics such as Fourier tra Nsform, fractional Fourier transform and other technologies and related technologies such as compressive sensing, computational ghost imaging and other methods to encrypt and decrypt encrypted information, hide and authenticate digital watermarking. Compared with traditional security systems, the advantages of optical and photon based security systems are light waves. More complex degrees of freedom such as strength, phase, polarization, nonlinear transformation and quantum properties of photons, in addition, the optical information can be reused, making the security system more secure and difficult to crack. This paper will first introduce the latest encryption and decryption and watermark technology at home and abroad, and systematically comb the common optical image encryption. On this basis, this paper will use the iterative phase recovery algorithm, combined with multiplexing technology and sparse data representation technology to develop the research work, and propose novel and feasible improvements. The specific research contents include the following aspects: (1) combining the iterative phase recovery algorithm with the pure phase mask multiplexing technology, the multiple images are entered. Using the same set of encrypted images to decrypt multiple images at the same time, it can greatly improve the utilization of encrypted images. (2) combined with the iterative phase recovery algorithm, space location reuse and sparse data representation, the space position of the phase part of the transformation plane calculated by the iterative phase recovery algorithm is relocated. The results show that the decrypted images obtained through space location multiplexing can not be resolved directly through the naked eye or the correlation coefficient, but can be resolved through the nonlinear correlation coefficient, which greatly reduces the amount of data processing and transmission, and greatly improves the security of the system. (3) (3) a cascade multi amplitude image coding and authentication scheme based on the theory of information optics is proposed. This paper designs and implements an optical hierarchical multi amplitude image coding and authentication scheme by using iterative phase recovery algorithm, pure phase mask multiplexing, spatial location reuse and sparse data representation. The first stage authentication image is coded as the complex amplitude field, then the amplitude part of the amplitude part is sampled and the amplitude part of the composite image is obtained. The amplitude of the composite image is limited by the output amplitude. The phase recovery algorithm is used to encode the second level authenticated image as a pure phase mask and can be obtained by the pure phase mask multiplexing technique. Finally, the final pure phase ciphertext is further encoded by the iterative phase recovery algorithm. Finally, the decryption key of the two level image authentication is designed by a specific formula. In the authentication process, only the corresponding secret key is placed in the corresponding position, and the nonlinear correlation distribution and the correlation system are used. On the basis of common optical multi amplitude image coding and authentication, the scheme not only increases the number of coded images, but also achieves hierarchical authentication on the basis of common optical multi amplitude image coding and authentication.

【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP309.7

【相似文献】