Deep Convolutional Neural Networks-Based Machine Vision System for Detecting Tomato Leaf Disease

Immediate identification of plant disease is one of the important solutions in Agricultural problems. In this study, the researchers develop an early detection system for tomato leaf diseases. It is important to create a system that will detect and classify a certain disease present in the leaf to prevent further loss. In order to do that, the researchers used an algorithm called YOLOv3 for training a model that accurately detects specific diseases for tomato leaves. The proposed model is able to classify the diseases and has a mean average precision(mAP) of 98.28 %. The result of the trained model varied with the accuracies ranging from 75% - 99%, for detecting the two common tomato leaf diseases such as, Early Blight and Septoria Leaf Spot.