Detection of plant diseases via computer vision based systems are being used to identify plant diseases promptly, to prevent the spread of the disease. In this work, we present a system for classifying plant diseases from photographs of the diseased parts of the plant. The system is trained using transfer learning on convolutional neural networks (VGGNet) which are trained to classify 38 plant diseases or 11 disease classes. 95.09% average accuracy was obtained on plant-disease classification using PlantVillage dataset. On the other hand, 96.27% average accuracy was obtained for disease classification instead of plant-disease classification. VGGNet obtained these scores.
