Tomato Leaf Disease Classification using Convolutional Neural Network Model

Abstract

Tomatoes and potatoes are the most commonly grown crops worldwide. The tomato contains antioxidant properties to prevent the cells from radicals. It can also protect the human body from cancer and sunburn. Tomatoes come under the human essential diet. But there are various diseases that are harmful to the plant's health and have an impact on its growth. Accurate and rapid disease identification of tomato plants is critical to increasing their long-term agricultural productivity. Disease identification can be done on the basis of physical changes in the leaves of the tomato plant. Detection of plant leaf diseases at early stages is beneficial to the Indian economy. In this work, tomato leaf diseases have been detected and classified using a deep learning model called Convolutional Neural Network. For performing the implementation, the Kaggle-based tomato leaf dataset has been used. This dataset contains nine different kinds of leaf diseases and a healthy leaf class. The proposed CNN model has been implemented with Adam and SGD optimizer. The results have shown that SGD performs best when compared with Adam. The accuracy and loss metrics have been used for predicting the performance outcome of the classification model. The results show that the CNN model achieves an accuracy (0.9966) and loss value(0.0044) when implemented with SGD optimizer.