Title: Detection and Classification of Tomato Leaf Diseases Using Deep Learning
Authors: Mazen S. M. Ihlayyel, Samy S. Abu Naser
Volume: 9
Issue: 6
Pages: 21-28
Publication Date: 2025/06/28
Abstract:
Leaf diseases significantly endanger tomato crop production and profitability, as tomatoes are widely grown worldwide. Early detection is critical to avoid substantial losses. We propose a new deep learning-based method for disease detection and prediction in tomato leaves. Convolutional Neural Networks (CNNs) are recognized as the most effective deep learning algorithm for image classification. Leveraging CNNs, we employed a CNN architecture to detect and identify diseases in tomato leaf images. The aptitude of CNNs for detection and prediction makes them ideal for this study. Our dataset included 1500 images of healthy and diseased tomato plants from the PlantVillage dataset. Training our CNN model on this dataset yielded a promising test accuracy of 92.93%. This high accuracy demonstrates the efficacy of our approach in accurately predicting diseases in tomato leaves. Our study aims to facilitate the early detection and prevention of leaf diseases in tomatoes, thereby improving crop yields and ensuring the profitability of tomato cultivation.