Plant Disease Diagnosis and Classification using Deep Learning

Abstract

Plant disease diagnosis and classification is an important task in agriculture as it helps in early identification and control of plant diseases, ultimately reducing crop loss and improving food security. In recent years, with the improvements in com-puter vision and machine learning, researchers have developed various techniques for the automated characterization and identification of plant diseases using images. This paper provides an overview of modern technology used for plant disease diagnosis and classification, including classification, image preprocessing, and feature extraction methods. Additionally, this paper highlights the challenges and future directions in this field, such as improving the accuracy of disease detection. In this paper, the model is trained and tested using a convolutional neural network (CNN). The Plant Village dataset is used to determine the precision of the model. In this paper, 18275 images were taken from the dataset belonging to 17 different classes. The dataset is split into three folders – train, validate and test. Images are distributed as follows: 70% are supplied to the train folder, 10% to the validate folder, and 20% to the test folder.