Automated Detection and Alerting System for Wheat Leaf Diseases Using the VGG16 Deep Learning Model

The rapid detection and classification of wheat leaf diseases are critical for ensuring crop health and maximizing yield. Conventional approaches to disease identification are laborious and frequently imprecise, resulting in substantial financial setbacks. This study presents a system that uses the VGG16 deep learning model to automatically detect and notify about diseases in wheat leaves. The system utilizes high-quality image datasets obtained from Kaggle and applies normalization, scaling, and augmentation approaches to improve the model's resilience. By utilizing the convolutional and pooling layers of VGG16, the system efficiently pulls complex features from the images, enabling precise disease categorization. The model training process is enhanced by finetuning the hyperparameters, resulting in exceptional performance metrics: 98.5% accuracy, 97.2% precision, 96.8% recall, and a 97.0% F1-score. These metrics demonstrate a considerable improvement over the performance of previous models. This technological innovation offers a pragmatic instrument for farmers, facilitating early identification of diseases and enhancing crop administration. Our research showcases improved precision and dependability, while also highlighting the possibility of incorporating sophisticated deep learning methods in agricultural applications. This opens up opportunities for future studies to broaden the range of datasets and investigate alternative architectures to achieve even better results.