POTATO LEAF DISEASE DIAGNOSIS SYSTEM

Abstract

The project Potato Leaf Disease Diagnosis system is used for detection and identification of disease related to potato plant. It is based on deep learning model using convolution neural network. In the agriculture system, one of the major problems in the plant is its disease. The naked eye observation of leaf is the traditional approach adopted in practice for detection of plant disease. Mostly diseases in plants are seen on the leaves. So it is possible to detect disease by the analysis of leaves of plant. Proper identification of disease on crop health can facilitate the control of disease through proper management strategies. A multilayer neural network with CNN is developed to train the system. The various steps followed for CNN model are convolution, max pooling, flattening and full connection. The system is trained with large number of images of potato leaves. After completing a training process a model is generated. This generated model is further used for data processing operation. In this system the user need to upload the image of potato leaf and the processing is performed using the train model and system detects and predicts whether the plant is healthy or not. We will train the system for predicting early blight, late blight, healthy plant and leaf mosaic condition. Based on the prediction the system provides information about the cause of disease and its treatment methods.