Maize Leaf Diseases Identification using Data Augmentation and Convolutional Neural Network

The economy of a country highly relies on agricultural productivity. This is the major reason that disease detection of crop plants plays an important part in the agricultural field. Plant leaves are affected by a wide range of diseases. In today’s scenario identification of different types of diseases is difficult for inexperienced farmers. The manual identification of diseases requires experts and continuous monitoring of fields for a long time. Identifying the diseases early can help to take appropriate measures quickly. Hence automatic approaches will help the farmers easily identify the diseases. This paper aims to improve the accuracy for the identification of maize leaf diseases by using Deep Learning for Convolutional Neural Network (CNN). The GoogleNet CNN architecture is used over the dataset randomly collected from Google for identifying nine kinds of maize leaf diseases. The highest accuracy obtained for the model is 99.87% and 98.55% for training and testing sets respectively.