Tomato Leaf Disease Classification using VGG16

This study proposes a novel image analysis model to detect the presence of infections in tomato crops. In order to categorize the disease, the tomato leaf images scraped from the Plant Village dataset, a convolutional neural network (CNN) model was developed. Among the numerous measures used to assess the performance of the model were training, validation and testing efficiencies as well as the number of trainable parameters in total. The research indicates that the Xception model had the best performance of 99%, VGG16 99%, InceptionV3 98% and ResNet50 95%. This method may assist farmers in recognizing diseases sooner and more precisely, thus improving the quality of the produced crops, the harvests and profitability in general. The results of the trial have demonstrated the efficacy of Xception and VGG16 in image classification with 99% accuracy for the classification tasks. When compared to ResNet50 and Inception V3, Xception and VGG16 provided substantial performance gains, suggesting that the more complicated and deeper structure of Xception and VGG16 improves their ability to extract features from tomato leaf diseases. The results of this study present a viable way forward for resolving the problems associated with tomato disease detection in agricultural environments. Through the facilitation of prompt detection and remediation, this technology has the potential to optimize crop management, minimize financial losses, and augment food security. To further optimize agricultural operations, future study could investigate how to integrate this system with automated disease control techniques and real-time monitoring. This paper aims to meet the goal for industry, innovation and infrastructure.