Chaotic colour image encryption based on differential evolutionary deep learning
Zhengbao Cai · International Journal of Information and Communication Technology · 2024
Traditional chaotic image encryption methods have certain limitations in terms of processing high-dimensional data, computational complexity and attack resistance, which limits their widespread promotion and use in practical applications. To solve these problems, this paper proposes a chaotic colour image encryption method based on differential evolutionary deep learning. Firstly, the security and stability of the image encryption algorithm is enhanced by introducing a six-dimensional cellular neural network (CNN). Secondly, the parameters of the six-dimensional CNN are optimised using differential evolutionary algorithms to improve the complexity and randomness of the chaotic sequences. The experimental results show that compared with the traditional CNN, AES and Chaotic Encryption Algorithm, this paper's method shows significant improvement in NPCR and UACI indicators.