Oil palm tree detection and health assessment with a machine learning model using UAV imagery

Tamilarasu, Tarrshan and Mohd Yusoh, Zeratul Izzah and Syed Ahmad, Sharifah Sakinah and Jason, Nathanieal Noah Immanual (2025) Oil palm tree detection and health assessment with a machine learning model using UAV imagery. International Journal of Artificial Intelligence and Expert Systems, 14 (2). pp. 26-36. ISSN 2180-124X

Text 00485151020251246182314.pdf Download (1MB)

Abstract

The oil palm industry is a vital contributor to Malaysia’s economy but continues to face challenges in plantation monitoring and productivity management. Conventional approaches for tree counting and health assessment are labour-intensive, costly, and prone to errors. This study introduces a machine learning approach utilizing unmanned aerial vehicle (UAV) imagery to automate the detection and classification of oil palm tree health. UAV imagery was processed into a two dimensional orthomosaic map, which was analysed with trained deep learning models capable of identifying individual trees and categorising their health status. The results achieved an overall F1-score of 0.84, with healthy trees classified most accurately. Tree counting accuracy exceeded 90%, and precision–recall analysis demonstrated that the model maintained high precision at strong confidence thresholds, though threshold adjustments are required to optimise recall. Overall, the proposed model can reduce manual effort, time, and cost, while improving consistency in plantation monitoring. These findings highlight the potential of UAV-based digital agriculture solutions to support sustainable and data-driven oil palm management in Malaysia.

Item Type:	Article
Uncontrolled Keywords:	Unmanned Aerial Vehicles (UAVs), Machine Learning, Oil Palm Plantations, Tree Detection, Tree Health Assessment.
Divisions:	Faculty of Information and Communication Technology
Depositing User:	Norfaradilla Idayu Ab. Ghafar
Date Deposited:	11 Aug 2026 03:37
Last Modified:	11 Aug 2026 03:37
URI:	http://eprints.utem.edu.my/id/eprint/30285
Statistic Details:	View Download Statistic

Actions (login required)

View Item