Disease Detection on Oil Palm Tree UAV Image Using Machine Learning

Tee, Zi Choong (2024) Disease Detection on Oil Palm Tree UAV Image Using Machine Learning. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text TEE ZI CHOONG_Full Text.pdf Restricted to Registered users only Download (5MB)

Abstract

As of September 2021, Malaysia ranked among the top producers of palm oil globally, second only to Indonesia. The palm oil industry is vital to Malaysia’s economy, highlighting the critical need for early detection of Basal Stem Rot (BSR) disease, which poses a significant threat to crop health and production. This study addresses this pressing concern by employing unmanned aerial vehicles (UAVs) to capture RGB images, subsequently transforming them into Hyperspectral Imaging (HSI) and greyscale representations. The aim of this research is to enhance the precision of BSR detection through Support Vector Machine (SVM) classification. While initially exploring alternative methods such as K-Nearest Neighbors (KNN), Random Forest (RF), and Decision Trees (DT), SVM emerges as the preferred choice due to its superior accuracy and efficiency. By leveraging HSI images and SVM classification, this study offers a detailed and effective approach to identifying early signs of BSR. This research constitutes a significant contribution to the advancement of sustainable agriculture within Malaysia’s palm oil industry, providing timely insights for proactive intervention and targeted disease management.