基于离散Hopfield网络的混沌图像加密算法  

Chaotic Image Encryption Algorithm Based on Discrete Hopfield Network

作　　者：刘祝华[1] 曾高荣[1] 谢芳森[1] 

机构地区：[1]江西师范大学物理与通信电子学院,南昌330022

出　　处：《计算机工程》2012年第4期112-115,共4页Computer Engineering

基　　金：江西师范大学青年基金资助项目(2704);江西师范大学博士基金资助项目(3338)

摘　　要：混沌图像加密算法在进行图像像素值置乱时,大多采用像素值整体处理的方式,因此难以抵抗已知明文攻击。针对该情况,提出一种基于离散Hopfield网络的高维混沌图像加密算法。使用Rssler三维混沌序列中的两维动态调整所有像素每个比特位的置乱权值及阈值,实现图像像素值置乱,剩下的一维用于图像像素位置置乱,由此提高破译难度。实验结果表明,该算法可抵抗差分攻击和统计分析,鲁棒性较强。For most of chaotic image encryption algorithms, pixel overall treatment method（such as XOR operation） is used in the image pixel value scrambling. It was very weak to resist known cleartext attack. So a high dimension chaotic encryption based on discrete Hopfield network is proposed in this paper. Image pixel value scrambling is realized through dynamic adjustment of each bit＇s scrambling weight and threshold depending on any two dimensions of R6ssler 3D chaotic sequences. Image pixel position scrambling is realized depending on the third dimension of Rossler 3D chaotic sequences. Each pixel value scrambling is a composite treatment, it increases decryption difficulty. Experimental results show that the algorithm has strong ability to resist attacks, and is a security image encryption algorithm.

关 键 词：离散HOPFIELD网络 混沌 Rssler系统 图像加密 置乱 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]