Deteksi Jenis Penyakit Tanaman Tebu Berdasarkan Citra Daun Menggunakan SSD MobileNet V2

Abstract

Sugarcane, scientifically known as Saccharum Officinarum Linn, is a plant used as the raw material for sugar production. However, the productivity of sugarcane growth continues to decline, as reflected in the low sugar content in the sugarcane stalks. One of the factors affecting sugarcane quality is the presence of pests and diseases on the sugarcane leaves. Currently, leaf disease detection in sugarcane plants is still performed manually through human visual observation, which is time-consuming and less accurate. To enable faster and more accurate detection, a digital image processing system is required. In this study, sugarcane leaf disease detection is conducted using the SSD MobileNet V2 method. The detection focuses on three types of sugarcane leaf diseases: rust spots, yellow spots, and ring spots. Based on the experiments conducted, the system developed using SSD MobileNet V2 achieved a detection accuracy of 88,29%.