基于新型混沌细胞神经网络的图像加密算法  

Image Encryption Algorithm Based on Chaotic Cellular Neural Network

作　　者：马英杰[1] 郑子怡 赵耿[1] 曾萍[1] MA Ying-jie;ZHENG Zi-yi;ZHAO Geng;ZENG Ping(Beijing Electronic Science and Technology Institute,Beijing 100070,China)

机构地区：[1]北京电子科技学院,北京100070

出　　处：《计算机仿真》2022年第12期271-274,共4页Computer Simulation

基　　金：国家自然科学基金项目(61772047);北京高校“高精尖”学科建设项目(3201017)。

摘　　要：一般性的流密码对图像加密难以抵抗差分攻击,固定的密钥使得算法对明文不具有敏感性,导致安全性大大降低。针对上述问题,提出一种基于新型混沌细胞神经网络的图像加密算法。构造基于符号函数的新型全互联四阶细胞神经网络,基于生成的混沌序列,根据明文像素点异或的结果选取密钥对图像加密,保证算法对于明文具有足够的敏感性。使用位置置乱和像素值替代两种方法加密图像,得到最终的密文。仿真结果表明,提出的算法能够有效抵御差分攻击,具有高度密钥敏感性,明文敏感性以及相邻像素间近似于零的相关性,保证了算法的安全性和可靠性。Using general stream cipher to encrypt images will be difficult to resist differential attacks, and the fixed keys will make the algorithm insensitive to plaintext, which will result in greatly reduced security. In order to solve this problem, an image encryption algorithm based on a new chaotic cellular neural network is proposed. A new type of fully interconnected fourth-order cellular neural network based on sign function is proposed. Based on the generated chaotic sequence, the key is selected to encrypt the image according to the XOR result of plaintext pixel points to ensure that the algorithm has sufficient sensitivity to plaintext. The final ciphertext is obtained by using position scrambling and pixel value substitution to encrypt the image. The simulation results show that the proposed algorithm can effectively resist differential attacks, has high key sensitivity, and plaintext sensitivity and the correlation between adjacent pixels is close to zero, which guarantees the security and reliability of the algorithm.

关 键 词：混沌细胞神经网络 图像加密 符号函数 明文敏感性 抗差分攻击 

分 类 号：TN918[电子电信—通信与信息系统]