摘要
为了更加系统了解基于混沌神经网络的图像加密技术,进行了简单研究。神经网络实现了混沌和扩散函数,利用混沌映射作为神经层的传递函数,设计在同一加密周期内对大量相邻图像像素进行混淆和扩散的CNN网络。其中,混沌神经元层实现数据扩散,线性神经元层实现数据混淆,两层重复多次增强密码,且解密过程与加密过程对称。与现有图像加密方法相比,该加密方法旨在减少相邻图像像素之间的冗余,获得更高的感知安全性。理论分析和实验结果表明,分组密码具有良好的计算安全性,更适合于图像加密。
In order to understand the image encryption technology based on chaotic neural network more systematically, a simple study was carried out. Chaos and diffusion function are realized by the neural network. Using chaotic mapping as the transfer function of the neural layer, a CNN network is designed to confuse and diffuse a large number of adjacent image pixels in the same encryption cycle. Among them, the chaotic neuron layer achieves data diffusion, the linear neuron layer achieves data confusion, the two layers repeatedly enhance the cipher, and the decryption process is symmetrical with the encryption process. Compared with the existing image encryption methods, this encryption method aims to reduce the redundancy between adjacent image pixels and achieve higher perception security. Theoretical analysis and experimental results show that block cipher has good computational security and is more suitable for image encryption.
作者
邹鹏
刘晋宏
蔡必汉
Zou Peng;Liu Jinhong;Cai Bihan(School of Computer Science and Information Engineering, Hubei University, Wuhan Hubei 430062, China)
出处
《信息与电脑》
2019年第15期48-49,52,共3页
Information & Computer
基金
湖北省大学生创新训练项目基金（项目编号：201810512051、201710512051）
关键词
混沌
神经网络
信息安全
图像加密
chaos
neural network
information security
image encryption
