基于Hopfield混沌神经网络的彩色图像加密算法研究

Color Image Encryption Algorithm based on Hopfield Chaotic Neural Networks

本文提出了一种基于混沌控制参数的置乱扩散结构的图像加密算法。在本算法中,复合混沌映射被用于生成置乱阶段的控制参数,用以进行图像洗牌,置乱图像像素之间的高度相关性。在扩散阶段,不同初始条件和控制参数的复合混沌映射作为生成Hopfield混沌神经网络的初始状态,迭代该神经网络,生成扩散阶段的密钥流。对算法进行仿真实验及数值统计分析表明,本算法具备良好的抗攻击性能,达到相应的安全水平。

This paper put forward an image encryption algorithm based on the parameters of chaos control for permutation-diffusion structure.In this algorithm,Tent-Logistic map is used to generate permutation phase of the control parameters,analyze high correlation between shuffle image and permutation image pixels.In the diffusion stage,Tent-Logistic map for different initial conditions and control parameters generate initial state of Hopfield chaotic neural network,iterate that neural network,and generate the key stream for diffusion stage.Algorithm simulation experiment and numerical statistics analysis shows that the algorithm has good performance against the attack,reach the corresponding safety level.