Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection

Luo, Yang; P. P. Abdul Majeed, Anwar; Omar, Zaid; Aslam, Saad; Chen, Yi

doi:10.3390/technologies13060234

Open AccessArticle

Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection

by

Yang Luo

¹

,

Anwar P. P. Abdul Majeed

^2,*

,

Zaid Omar

³

,

Saad Aslam

² and

Yi Chen

^1,*

¹

School of Intelligent Manufacturing Ecosystem, Xi’an Jiaotong-Liverpool University, Taicang, Suzhou 215400, China

²

School of Computing and Artificial Intelligence, Faculty of Engineering and Technology, Sunway University, Bandar Sunway, Subang Jaya 47500, Malaysia

³

Faculty of Electrical Engineering, Universiti Teknologi Malaysia, Johor Bahru 81310, Malaysia

^*

Authors to whom correspondence should be addressed.

Technologies 2025, 13(6), 234; https://doi.org/10.3390/technologies13060234

Submission received: 14 April 2025 / Revised: 30 May 2025 / Accepted: 4 June 2025 / Published: 6 June 2025

(This article belongs to the Section Manufacturing Technology)

Download

Browse Figures

Versions Notes

Abstract

The present study addresses the inefficiencies of the manual classification of oil palm fresh fruit bunches (FFBs) by introducing a computationally efficient alternative to traditional deep learning approaches that require extensive retraining and large datasets. Using feature-based transfer learning, where pre-trained Convolutional Neural Network architectures, namely EfficientNet_B0, EfficientNet_B4, ResNet152, and VGG16, serve as fixed feature extractors coupled with the Logistic Regression classifier, this research evaluated the performance on a dataset of 466 images categorized as defective or non-defective. The results demonstrate a robust classification performance across all architectures, with the EfficientNet_B4–LR pipeline achieving an exceptional accuracy value of 96.81%, which was further enhanced through hyperparameter optimization. This confirms that feature-based transfer learning offers a reliable, resource-efficient, and practical solution for automated FFB defect detection that can significantly benefit the palm oil industry by providing a scalable alternative to subjective manual-grading methods.

Keywords: fresh fruit bunch; defect detection; transfer learning; feature extraction; deep learning

Share and Cite

MDPI and ACS Style

Luo, Y.; P. P. Abdul Majeed, A.; Omar, Z.; Aslam, S.; Chen, Y. Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection. Technologies 2025, 13, 234. https://doi.org/10.3390/technologies13060234

AMA Style

Luo Y, P. P. Abdul Majeed A, Omar Z, Aslam S, Chen Y. Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection. Technologies. 2025; 13(6):234. https://doi.org/10.3390/technologies13060234

Chicago/Turabian Style

Luo, Yang, Anwar P. P. Abdul Majeed, Zaid Omar, Saad Aslam, and Yi Chen. 2025. "Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection" Technologies 13, no. 6: 234. https://doi.org/10.3390/technologies13060234

APA Style

Luo, Y., P. P. Abdul Majeed, A., Omar, Z., Aslam, S., & Chen, Y. (2025). Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection. Technologies, 13(6), 234. https://doi.org/10.3390/technologies13060234

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

Article Menu

Computationally Efficient Transfer Learning Pipeline for Oil Palm Fresh Fruit Bunch Defect Detection

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI