Klasifikasi Penyakit Daun pada Tanaman Kopi dengan Penerapan Metode Faster Region Convolutional Neural Network (Faster R-CNN)

Abstract

Coffee is a high-value agricultural commodity with a significant contribution to foreign exchange for the Indonesian state. With a cultivation area of 1.24 million hectares, Indonesia is among the largest coffee producers in the world. Despite this substantial potential, it has not been fully realized due to the ineffective management of disease prevention and control. Diseases can be indicated by changes in the shape and color of the leaves. However, factors such as vision, experience, as well as the vast complexity of the land and the large number of coffee plants pose challenges for farmers. Research is conducted to find an alternative solution by developing a computer vision-based system using the Faster R-CNN application to classify three types of leaf diseases: Leaf Rust, Leaf Blight, and Leaf Miner. This study employs a dataset comprising 3,600 images, divided into 2,880 training data, 360 validation data, and 360 testing data. Testing results of the system implementing the Faster R-CNN method achieve an accuracy value of 95%.