Design and Implementation of Deep Learning Models for Tomato Plant Leaf Disease Classification

Agriculture is an integral part of the community and making sure that plants can grow and produce fruits and vegetables is necessary for the survival of the human race. Here we look into the classification of various tomato plant diseases such as early blight, Septoria leaf spot, bacterial leaf spot, late blight, mosaic virus, yellow virus, mold leaf, and two spotted spider mite leaf. We have selected tomato plants as they are the world's most used and common fruit as it is used in every household throughout the world. The dataset being used is a combination of various datasets such as the PlantVillage dataset and the PlantDoc Dataset along with augmentation. The dataset has been taken from Kaggle and contains more than 40,000 images of tomato plant leaves. Various deep learning models such as the pre-trained (such as ResNet50, ResNet152, EfficientNetB3, Xception, VGG19, and InceptionV3) and a suggested Convolutional Neural Network model are all compared to see which provides the best results.