Plant Disease Classification using Ensemble Deep Learning

It is essential to predict plant diseases in agriculture and combat them promptly. Thus, with early intervention, both the yield and the quality of the product can be increased. This will make a great financial contribution to the people engaged in agriculture. This study made plant disease classification using the PlantVillage dataset, which is open to access. First, features were obtained by using transfer learning in AlexNet, VGG16, and ResNet18 deep learning networks. Then, using the features obtained from these networks, classification results were obtained using nearest neighbor, support vector machines, and artificial neural network classifiers. Finally, the results obtained using the bagging method from ensemble learning algorithms were compared.