Arza Faidillah, Siti (2025) Deteksi Penyakit Pada Daun Kelengkeng Menggunakan Gabor Filter Dan K-nearest Neihbor(k-nn). Other thesis, Universitas Islam Riau.

Text 203510778.pdf - Submitted Version Restricted to Registered users only Download (3MB) | Request a copy

Abstract

The longan plant (Dimocarpus longan) is a high-value fruit commodity, but is often threatened by various diseases that can reduce productivity and product quality. Early detection of diseases on longan leavesis very important to maintain sustainable production. This research aims to develop a disease detection system on longan leaves using the Gabor Filter and K-Nearest Neighbor (K-NN) methods. Themethod used included collecting 152 longan leaf image data, which was then augmented into 1064 images. The feature extraction process is carried out using CIELAB color space for color analysis and Gabor Filter for texture analysis. Classification was carried out using the K-NN algorithm. The research results showed that the system was able to identify types of disease on longan leaves, namely Anthracnose, Cercospora, and healthy leaves. Testing shows that the best K value for the K-NN algorithm is K = 1 with an accuracy of 76%. Model evaluation metrics show an average F1 score, precision, and recall of 76%. The conclusion of this research is that the system developed is effective in detecting diseases on longan leaves, although there is still room for further development, such as adding types of disease that can be detected and using other classification methods.

Actions (login required)

View Item