【作者】 张坤；

【导师】 肖华勇；

【作者基本信息】 西北工业大学 ， 统计学， 2006， 硕士

【摘要】 计算机技术水平的迅速提高与现代通信技术的广泛普及,为信息隐藏的发展提供了更加广阔的空间。信息隐藏是将秘密信息隐藏于一般的非秘密的数字媒体文件(如图像、声音、文档文件)中,从而不让非法者发觉的一种方法。针对数字图像信息隐藏的理论与算法,论文简要阐述了信息隐藏的基本知识和基础理论,并从以下三个方面进行深入研究: (1) 数字图像置乱算法研究 首先总结图像置乱算法的研究概况,研究了基于矩阵变换的图像置乱算法;针对AES的S盒特点,提出了基于AES的S盒图像置乱算法。通过实验仿真,运用灰度直方图、期望、标准差、相关系数等统计方法对图像置乱效果进行综合评价,验证了算法的有效性。 (2) 基于DCT域的图像信息隐藏算法研究 首先介绍DCT变换及其在图像信息隐藏中的应用,然后提出以下三种图像信息隐藏算法:基于模糊熵和自适应量化的图像信息隐藏算法;基于变异程度的图像信息隐藏算法;基于模糊熵的图像融合隐藏算法。算法中根据模糊熵、变异程度来选择嵌入子块及嵌入强度,从而在保证图像透明性的前提下,进一步提高算法的鲁棒性。通过实验仿真,运用峰值信噪比、相似性系数等分析了算法的抗攻击性能,结果表明以上算法具有较好的透明性、鲁棒性、实用性,是性能优良的图像信息隐藏算法。 (3) 基于DWT域的图像信息隐藏算法研究 简要介绍了离散小波变换及其在图像信息隐藏中的应用,提出一种基于纠错编码的图像信息隐藏算法。算法中首先对秘密图像进行纠错编码;然后根据小波系数特点选择嵌入系数及嵌入强度:最后在提取时,采用三种提取方法来优化提取效果,进一步提高了秘密信息恢复的效果。实验仿真与性能分析的结果表明,该算法具有较好的透明性、鲁棒性、实用性。更多还原

【Abstract】 The rapid enhancement of the computer technique and the widespread popularization of the communication technology are providing the broad space for the information hiding in research and application. The information hiding technique hides the secret information in the ordinary information (carrier information) in the process of transmission so as to guarantee the secret information not easily stolen or destructed. This dissertation details foundational knowledge of the technique and further analyzes the algorithms for information hiding in digital image, the main tasks are as follows:(1) The algorithms for image scramblingThe dissertation outlines the research situation of image scrambling, and then studies one algorithm based on matrix transformation and proposes one algorithm based on S box in AES. In order to confirm the validity of the algorithm, the performances of scrambling image are analyzed by the gray histogram, the expectation, the standard deviation, the correlation coefficient etc.(2) The algorithms for information hiding in digital image based on DCT The dissertation introduces DCT and its application for the information hiding indigital image. And three algorithms are proposed, one algorithm based on fuzzy entropy and auto-adapted quantification, one algorithm based on variation degree, one algorithm based on fuzzy entropy and image fuzzy. The experimental results show that the proposed algorithms are perceptually invisible and robust to some image processing such as loss JPEG compression, cropping, and noise.(3) The algorithm for information hiding in digital image based on DWT One kind of blind information hiding in digital image based on error correctingcoding and the DWT characters is proposed. Experimental results indicate the algorithm has a better transparency, robustness and usability.更多还原