A chaotic two-dimensional image classification algorithm based on convolutional neural networks

Zhou Xuefang, Wang Hongliang, Hu Junchao, Haozhen Li, Mengmeng Xu and Hu Miao

Chaos, Solitons & Fractals, 2025, vol. 195, issue C

Abstract: To evaluate deep learning's classification capabilities on extensive datasets and its noise tolerance, a chaotic two-dimensional (2D) image classification algorithm using a convolutional neural network is introduced. 1D time series from circuit and laser chaotic systems are transformed into 128 × 128 grayscale images to create a diverse dataset. A four-layer convolutional neural network (CNN) then classifies seven types of chaos, then results show that the system achieves an accurate chaos classification rate of up to 99.56 %. This paper also examines the impact of network depth, noise intensity, and image pixel size on performance. With a four-layer network, the system shows higher accuracy and lower performance loss. Adding Gaussian noise with a variance of 0.1 still maintains accuracy above 98.5 %. Increasing image resolution to 128 × 128 pixels results in a stable accuracy of 99.3 %. Overall, the system demonstrates strong robustness and generalization ability.

Keywords: Deep learning; Chaotic sequences; Chaotic classification; 2D images (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0960077925003042
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:chsofr:v:195:y:2025:i:c:s0960077925003042

DOI: 10.1016/j.chaos.2025.116291

Access Statistics for this article

Chaos, Solitons & Fractals is currently edited by Stefano Boccaletti and Stelios Bekiros

More articles in Chaos, Solitons & Fractals from Elsevier
Bibliographic data for series maintained by Thayer, Thomas R. ().