Novel chaotic colour image cryptosystem with deep learning

A chaotic colour image cryptosystem frame with deep learning is proposed in this paper. First, we choose a chaotic system, after which a long short-term memory network is used to train 4D hyper-chaotic Lorenz signals and forecast four new signals, which passed the randomness test. Finally, we applie...

Full description

Saved in:
Bibliographic Details
Published in:Chaos, solitons and fractals Vol. 161; p. 112380
Main Authors: Zhou, Shuang, Zhao, Zhipeng, Wang, Xingyuan
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.08.2022
Subjects:
Chaotic system
Deep learning
Image encryption
Deep learning
Chaotic system
Image encryption
ISSN:0960-0779, 1873-2887
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A chaotic colour image cryptosystem frame with deep learning is proposed in this paper. First, we choose a chaotic system, after which a long short-term memory network is used to train 4D hyper-chaotic Lorenz signals and forecast four new signals, which passed the randomness test. Finally, we applied them to the colour cryptosystem. Given that new chaotic signals generated by deep learning are different from the original chaotic signals and the complex structure of deep learning, it is difficult to be attacked. Simulation results indicate that the proposed method has a high level of security compared to the representative image encryption algorithms.
ISSN:0960-0779
1873-2887
DOI:10.1016/j.chaos.2022.112380