Maize Leaf Disease Detection Using Convolutional Neural Network

Crop diseases represent a big hazard to food security. The primary basis for recognizing plant diseases based on leaf colour is the colour information of diseased leaves. In the agricultural field, the identification of crop diseases is required with higher accuracy. We propose a profound convolutional neural organization (CNN)-based design for maize leaf sickness order. The test is completed with maize leaf pictures from the maize dataset. The proposed CNNs have been prepared to perceive four particular classes, three of which are sicknesses (common rust, grey leaf spot, and blight) and one of which is healthy. The Xception model has a 96.5% accuracy rating, while the Inception model has an 87.13% accuracy rating.