Classification of Tomato Leaf Disease Using a Custom Convolutional Neural Network

Abstract

A plant's genetic potential for crop production can only be realised if the plant is healthy. Infected plants produce less than their genetic potential when they are unhealthy and exposed to infection-causing agents of any kind. A disease can have an impact on a plant's metabolism. Manual checking of plant health is not feasible for anytime. Accurately identifying the disease as soon as it first manifests on the plant is crucial for controlling it in farms. Thus, taking the proper action to stop further crop and yield damage will depend heavily on an automated method of disease identification and precise disease relegation. This paper presents a convolutional neural network (CNN) model for diagnosing tomato leaf diseases. The findings are presented with an emphasis on accuracy as well as loss. About, 14240 numbers of tomato leaf image data representing nine distinct disease classes were utilized to train the model. On average, this classification was found to be 95.53 percent accurate.