Potato Plant Leaf Diseases Classification using Neural Network Models

Throughout the process of plant development, there is a possibility of encountering diseases that can negatively affect the production of crops. Additionally, farmers may experience challenges in terms of overall yield and income rates. However, numerous advantages can be obtained by early identification and classification of these diseases. The advancements in contemporary technology have proven advantageous in categorizing plant diseases. This report aims to introduce a system for classifying plant diseases, specifically emphasizing potato plant leaves. The classification system utilizes neural network-based techniques, specifically the Convolutional Neural Network (CNN) models. The model classifies potato leaves into three categories, namely healthy leaves, early blight, and late blight. This system utilizes two CNN models- AlexNet and GoogLeN et, to carry out the classification process. To assess AlexNet's and GoogLeNet's performance, images of potato leaves are obtained from the Kaggle dataset. The results of the experiment demonstrate that the classification accuracy of potato plant leaf diseases by the system is 96.4 % for the AlexNet model and 97.0% accuracy for the GoogLeNet model.