隐写模糊安全性测度及其优化嵌入算法研究

项目名称： 隐写模糊安全性测度及其优化嵌入算法研究

项目编号： No.61462046

项目类型： 地区科学基金项目

立项/批准年度： 2015

项目学科： 自动化技术、计算机技术

项目作者： 欧阳春娟

作者单位： 井冈山大学

项目金额： 45万元

中文摘要： 隐写安全性测度、高容量低失真隐写算法设计是隐写通信研究领域的两大关键问题。本项目首先根据隐写通信中存在的模糊不确定性，通过构建隐写载体数据和载密数据的高阶模糊统计模型，采用闵科夫斯基距离、Vague集相似度量、模糊超香肠几何覆盖来探索隐写模糊安全性测度问题，进而在其理论指导下，研究不同智能优化方法对STC模型下的各类隐写算法的优化改进。一方面，从优化统计特征、优化决策面、优化嵌入位置等单目标优化角度，采用混合蛙跳优化算法、极值动力学优化、差分优化来研究隐写过程中的可优化问题；另一方面，从嵌入容量、安全性等隐写设计指标的相互制约关系，探索多目标智能优化隐写问题的新框架和新算法。本项目的研究可丰富隐写安全性理论，为设计高容量低失真的安全隐写算法提供新思路。

中文关键词： 信息隐藏；隐写；安全性测度；智能优化

英文摘要： There are two key problems in the research of the steganography communication security field. One is to evaluate the security of the steganographic system. The other is to design the high-capacity and low-distortion steganographic algorithms. First, according to the indeterminacy of steganography communication, the cover data and the stego data are modeled as a high-order fuzzy statistical model. And the Minkowski distance, vague set similarity measure and fuzzy super sausage geometric coverage are adopted to research the fuzzy security measures for the steganographic system. Then, under the guidance of this theory, the project optimizes the various steganographies of STC model by using different intelligent optimization methods. On the one hand, the single objective optimization problems in the steganographic process, such as statistical features, decision hyperplane, embedding position are studied by using SFLA(Shuffled Frog Leaping Algorithm), EO(Extremal Optimization) and DE(Differential Evolution) etc. On the other hand, according to the inter restricted relationships of the steganographic design ndexes,such as embedding capacity, security,and etc. The new frameworks and new algorithms of the multi-objective optimization steganographic problem are explored in this project.In a word, our research can enrich the steganographic security theory, and provide a new way to design the high-capacity and low distortion secure steganographic algorithms.

英文关键词： information hiding;steganography;security measure;intelligent optimization