Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification

Ng, Kuan-Huei; Azaman, Mohd Ikmal Hafizi; Udos, Waldo; Khalid, Mohd Ramdhan Mohd; Mohd Bakri, Mohd Azwan; Lim, Kok-Sing

doi:10.3390/electronics15030671

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification

by

Kuan-Huei Ng

^1,2

,

Mohd Ikmal Hafizi Azaman

^1,2,3,

Waldo Udos

¹,

Mohd Ramdhan Mohd Khalid

³

,

Mohd Azwan Mohd Bakri

³

and

Kok-Sing Lim

^1,*

¹

Photonics Research Centre, University of Malaya, Kuala Lumpur 50603, Malaysia

²

Institute for Advanced Studies, University of Malaya, Kuala Lumpur 50603, Malaysia

³

Malaysian Palm Oil Board, 6, Persiaran Institusi, Bandar Baru Bangi 43000, Malaysia

^*

Author to whom correspondence should be addressed.

Electronics 2026, 15(3), 671; https://doi.org/10.3390/electronics15030671

Submission received: 11 December 2025 / Revised: 2 January 2026 / Accepted: 9 January 2026 / Published: 3 February 2026

(This article belongs to the Special Issue Trends and Challenges in Integrated Photonics)

Download Versions Notes

Abstract

In this work, we present a Physics-Informed Neural Network (PINN) framework for the classification of oil palm fresh fruit bunch (FFB) ripeness using RGB images. Unlike conventional Convolutional Neural Networks (CNNs) that learn solely from visual patterns, the proposed PINN integrates a physics-based index—derived from the red-to-green pixel intensity ratio—directly into the network architecture and loss function. This hybrid design embeds wavelength-dependent physical knowledge related to chlorophyll degradation during ripening, enabling the model to learn more robust and generalizable features even with limited and imbalanced training data. The PINN model achieves a peak accuracy of 0.73, outperforming the purely data-driven CNN baseline (0.68) by a margin of 5%. Overall, the PINN demonstrates superior performance in minority-class detection and maintains stable convergence under three different lighting conditions (different light spectra). These results highlight the effectiveness of integrating domain-specific physical insights into deep learning models, offering a promising pathway toward reliable, non-destructive, and automated ripeness assessment for agricultural applications.

Keywords: physics-informed neural network; colorimetric index; agricultural

Share and Cite

MDPI and ACS Style

Ng, K.-H.; Azaman, M.I.H.; Udos, W.; Khalid, M.R.M.; Mohd Bakri, M.A.; Lim, K.-S. Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification. Electronics 2026, 15, 671. https://doi.org/10.3390/electronics15030671

AMA Style

Ng K-H, Azaman MIH, Udos W, Khalid MRM, Mohd Bakri MA, Lim K-S. Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification. Electronics. 2026; 15(3):671. https://doi.org/10.3390/electronics15030671

Chicago/Turabian Style

Ng, Kuan-Huei, Mohd Ikmal Hafizi Azaman, Waldo Udos, Mohd Ramdhan Mohd Khalid, Mohd Azwan Mohd Bakri, and Kok-Sing Lim. 2026. "Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification" Electronics 15, no. 3: 671. https://doi.org/10.3390/electronics15030671

APA Style

Ng, K.-H., Azaman, M. I. H., Udos, W., Khalid, M. R. M., Mohd Bakri, M. A., & Lim, K.-S. (2026). Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification. Electronics, 15(3), 671. https://doi.org/10.3390/electronics15030671

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.

Article Menu

Physics-Informed Neural Network-Assisted Imaging for Oil Palm Fruit Ripeness Classification

Abstract

Share and Cite

Article Metrics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI