Classification of Leaf Diseases in Oil Palm Plants with Haar Wavelet Transform Features Based on Machine Learning

¹ Department of Electrical Engineering, Faculty of Engineering, Universitas Muhammadiyah Yogyakarta, 55183 Yogyakarta, Indonesia
² Faculty of Electrical Engineering and Technology, Universiti Malaysia Perlis, 02600 Perlis, Malaysia

^* Corresponding author: yjusman@umy.ac.id

Abstract

Oil palm plants are essential as they produce palm fruit that can be processed into edible oil—an essential human need. However, these plants are often infected with diseases, negatively impacting crop productivity and the quality of the oil produced. These diseases are caused by mushrooms, bacteria, viruses, and pests that can spread rapidly and damage the leaves. Therefore, early detection of oil palm leaf disease plays a crucial role in reducing the negative impact on crops and significant economic losses. This study aims to design a system to classify the types of leaf diseases of oil palm plants using texture feature extraction (Haar Wavelet Algorithm) and machine learning-based classification algorithms (Cubic SVM, Medium Gaussian SVM, Quadratic SVM, Cosine KNN, Fine KNN, and Weighted KNN). Cubic SVM yielded the highest training result with an averages accuracy of 81.54% and an average time of 48.135 seconds. However, Medium Gaussian SVM outperformed other models during testing, producing an accuracy of 87%, precision of 81%, recall of 81 %, specificity of 90%, and F-score of 81%.