Plant Disease Classification Using Deep Learning Framework

Abstract

This Agriculture is the main source of livelihood in India. Major destruction occurs in the field of agriculture is due to the diseases in plants. Disease causes heavy devastation in the field of farming. An automatic system needs to be developed for the prevention of crops at initial stages from diseases. The automatic system will diagnose plant disease and identify its category. The paper proposed an automatic disease detection system which recognize the category of the diseases and healthiness in plants. The classification of diseases have done on the basis of symptoms appeared on the leaf surface. The paper broadly categorises the diseases into three categories such as fungal, bacterial and viral. The deep learning approach namely Convolutional Neural Network is used for the classification of the diseases. The dataset consists of 64963 number of samples in which training has performed on 80% of the samples of dataset and validation has performed on 20% of the samples of dataset. The simulation results shows that the convolutional neural network that have been trained on given dataset acquired an accuracy of 99.12%.