Multiclass Disease Identification in Wheat Leaves Using Convolutional Neural Networks

The persistence of this study revolves around the application of deep learning techniques, specifically CNNs, for the classification of illnesses harmful to wheat. In view of this, the investigation is led by a database that consists of images of good wheat leaves and infected ones with stripe rust, as well as those affected by septoria. The program used Python for the building of a model and Jupyter as an interactive front-end. The two combined to form CNN, which then got trained with the running of fifteen epochs at a batch size of thirty-two to train for the best scores of measures such as accuracy, precision, recall, and F1-score. The experiment results show the CNN has correctly identified diseases in wheat, with a general accuracy of 96.00%. While this enhances the efficiency of the disease detection aspect in agriculture, it could have further promoted improvements in the crop management quality and ensure food security in place..