Classification of diseases in Paddy by using Deep transfer learning MobileNetV2 model

Abstract

Rice is a crop that is grown widely across the world. As the demand for healthy food is increasing nowadays and cultivators were very prone to the safe and demand for foods. The disease can affect the plants at any stage and causes more damage and loss to the cultivators. Identifying disease in an early stage can take necessary remedial actions in early stage and may reduce the damage caused. In this approach, we have classified nine different types of diseases which affect paddy. Recent developments in deep learning techniques were applied for getting better performance. In this paper, we have calculated the performance of MobileNetV2 which is one of the Deep transfer models. This model obtained an accuracy of 98% which is the best performance score. Along with this transfer learning model, we have developed our model which obtained a result of 98.73%. The time taken for training the model is less when compared with other models.