CNN Model Suitability Analysis for Prediction of Tomato Leaf Diseases

India’s most popular vegetable crop is tomatoes, but despite the tropical climate, tomato plant growth is affected by The circumstances of weather and other factors. Plant disease, along with natural calamities, is a major cause of economic losses due to their effects on agricultural productivity. The time and accuracy of conventional disease detection methods for tomatoes has long been unacceptable. Early disease monitoring, identification, and classification are crucial to addressing this issue, and deep learning has emerged as the preferred approach for automated object recognition and detection. Convolution Neural Networks (CNNs) have found widespread application in object detection, including the identification of tomato leaf disease. The literature ranks ResNet-101, VGGNet, AlexNet, LeNet, and Google Net as the top five CNN models for tomato leaf disease detection, based on a dataset of 10 distinct disease classes. VGGNet and ResNet-101 were implemented and analyzed, and it was observed that ResNet-101 achieved higher accuracy than VGGNet in most of the evaluated criteria.