Cotton Leaf Disease Prediction Using VGG16 and RESNET50

The farming business is a vital component of every economy, and its success is directly proportional to the yields of individual farmers. The micro-level illnesses that manifest themselves throughout the development of a plant often have an effect on the production of the plant under consideration. The presence of a cotton plant disease is a significant cause for worry because of the detrimental impact it has on crop output. Diagnostic procedures for these disorders need to be carried out on a consistent and accurate basis. Without this, there is a possibility that its output may suffer a major decline, which will eventually have an impact on the farmer economy. Diagnoses made manually take a significant amount of time. Consequently, this study is to make use of the advantages of deep learning to automate the diagnosis of disease in cotton plants. This study investigates the impact of deep learning on a selection of pre-trained models, including VGG-16 and ResNet50, and also proposes a hybrid method to predict cotton leaf disease. According to the examination of the results, the proposed model scored the greatest accuracy of 99.25% when compared to other CNN architectures that had been trained. Several criteria, including Precision, Recall, and F1 score, are used to measure the performance of the model.