结合整数规划的改进YASS算法研究-电子与通信工程专业论文.docxVIP

上海交通大学专业学位硕士论文摘要 上海交通大学专业学位硕士论文 摘要 万方数据 万方数据 结合整数规划的改进 YASS 算法研究 摘 要 随着信息技术的飞速发展，互联网逐渐渗透到人们生活的各个角落，成为日 常通信的主要渠道。信息隐藏作为一种无形的通信技术，通过将秘密信息嵌入常 见载体中以达到隐蔽通信的目的。JPEG（Joint Photographic Experts Group）格式 图像是一种在互联网上广泛应用的图片格式，也是一种被广泛研究的图像隐写载 体格式，如经典的 Jsteg，F5，OutGuess 等图像隐写算法都是基于 JPEG 图像。 同样是以 JPEG 格式图像为载体，YASS（Yet Another Steganographic Scheme）通 过在确定的图像分块中随机选取图像的子块进行 DCT（Discrete Cosine Transform） 变换和 QIM（Quantization Index Modulation）信息嵌入，可以有效抵抗主流的通 用隐写分析方法，表现出相当优秀的抗检测性能，引起了人们广泛的关注。 本文在深入研究了 YASS 算法的基础上提出一种改进的图像隐写算法。针对 YASS 隐写率不理想的问题，通过分析不同图像分块 DCT 系数间的联系，结合 整数规划，提出了一种迭代算法，在 YASS 隐写的基础之上对已隐写的图像载体 进行信息校正，能够有效减少在常规的 YASS 隐写算法过程中引入的误码，进而 减小整体 JPEG 图像压缩对嵌入信息的影响，大幅度提升隐写算法的信息嵌入率。 最后采用检测效率较高的 98 维马尔科夫通用隐写检测算法对隐写图像进行检测， 从理论和实验上都证明，所提方法在隐写嵌入率和抗检测能力方面都有很好的效 果。 关键词：YASS，整数规划，图像隐写，DCT I 上海交通大学专业学位硕士论文ABS 上海交通大学专业学位硕士论文 ABSTRACT THE RESEARCH OF IMPROVED YASS ALGORITHM COMBINED WITH INTEGER PROGRAMMING ABSTRACT With the rapid development of information technology, the Internet gradually penetrates into every corner of people’ lives and becomes the major channel of daily communication. As an invisible communication technology, information hiding embeds secret information into common hidden carriers in order to achieve concealed communication. JPEG image format is a picture format that widely used on the Internet and is also a widely studied image steganography carrier format. Such as Jsteg, F5 and OutGuess are based on JPEG images that are well known and classic. YASS – Yet Another Steganographic Scheme is also based on JPEG format image. It randomly selected sub-block in the determined image block and make DCT and QIM information embedding. YASS can effectively resist from popular universal steganalysis methods in which it shows excellent anti-detection performance and caused widespread concern. After in-depth study of YASS algorithm, the paper proposed a new scheme of steganography which is based on YASS. In order to resolve the imperfect steganographic efficiency, we discuss the re