CNNs in Crop Care: A Comparative Analysis of Tomato Disease Detection Models

Abstract

Timely and accurate detection of plant diseases is critical for sustainable crop production. This paper presents a comparative analysis of convolutional neural network (CNN) models for automated tomato leaf disease detection, including VGG16, VGG19, AlexNet and Google Net. A dataset of tomato leaf images with multiple disease types was utilized to train and test the performance of each CNN architecture. Our results demonstrate that the VGG16 model achieved the highest accuracy of 99.38% in classifying tomato leaf diseases. Compared to AlexNet and VGG19 models which obtained 98.67% and 99.29% accuracy respectively, the superior performance of VGG16 highlights its suitability for deploying robust tomato disease detection systems to enable prompt disease control interventions. Our work provides valuable insights on harnessing deep learning for agricultural disease surveillance, allowing farmers and agronomists to identify crop infections early and implement timely management practices to reduce yield losses.