基于分数阶时延混沌神经网络的图像加密

基于分数阶时延混沌神经网络,提出了一种新的图像加密算法:利用混沌系统产生秘钥流,把时延和分数阶导数嵌入到秘钥系统中以增加算法的安全性;用像素异或和置换相结合的方法对原始图像进行加密。通过对加密后图像各像素的水平垂直相关性、信息熵、游程统计、灰度变化平均值、直方图均衡度、密钥灵敏度等数据的详细分析,验证了该加密算法的安全性和有效性。

Image Encryption Based on Delayed Fractional-order Chaotic Neural Networks

In this paper,a new image encryption scheme was proposed based on a delayed fractional-order chaotic neu- ral networks. In the process of generating a key stream, the time delay and fractional derivative were embedded in the proposed scheme to improve the security. All pixels of original image were encrypted by combining method exclusive OR and replacement. This scheme was described in detail with security study including analysis of correlation, infor- mation entropy, run statistic, mean variance gray value, histogram equalization degree and key sensitivity. Experimen- tal results show that the newly proposed image encryption scheme possesses high security.