Pemanfaatan Yolo Untuk Deteksi Hama Dan Penyakit Pada Daun Cabai Menggunakan Metode Deep Learning

Abstract

Chili plants are one of the horticultural crops in Indonesia which have great potential in the Indonesian economy. However, crop failure often occurs. One of the main factors causing this is pest and disease attacks on chili plants. This requires early prevention which can reduce losses. With today's technological developments, prevention can be done easily and economically by using deep learning methods. YOLO is a deep learning algorithm that is commonly used to detect objects in real time. There are 4 classes that will be tested, namely leaves affected by yellow virus disease, leaf spot, thrips pests, and healthy chili leaves. Testing was carried out with a web-based application created with the flask framework. The accuracy results of the YOLO model training process with epoch 150 were 73%. The precision, recall and mAP values ​​obtained were 77.4%, 67.1% and 75.1%. Testing produces accuracy above 74%. The results of this research still produce accuracy that is not high enough, but the application can be used to detect it well and is quite accurate.