Sistem Pendeteksi Penyakit Daun Tanaman Jeruk Kintamani menggunakan Metode Naive Bayes dan Ekstraksi Fitur HOG

Abstract

Kintamani citrus is one of Bali’s native plants that offers many benefits, especially from its leaves, which are often used in traditional medicine. However, the potential of these leaves is often not fully utilized due to the lack of public awareness in recognizing disease symptoms on the leaves. Currently, the detection of diseases in Kintamani citrus leaves is mostly done manually, which can be less efficient and less accurate. This research aims to build a classification system for detecting diseases in Kintamani citrus leaves using digital images, by applying the Histogram of Oriented Gradients (HOG) for feature extraction and the Naive Bayes method for classification. A total of 100 leaf images collected from the field were used as the dataset. The process includes image preprocessing, feature extraction using HOG, and classification using Naive Bayes. The results show an accuracy of 90%, indicating that this method can be considered fairly effective, although there is still room for improvement. This research is expected to contribute to the development of a more efficient and automated system for detecting plant diseases, especially in the agricultural sector.