An Efficient Algorithm for Plant Disease Detection Using Deep Convolutional Networks

Abstract

Plant diseases and pests are important factors in determining crop yield and quality. Plant diseases are not only a threat to food security on a global scale but can also have devastating consequences for farmers whose livelihood depend on healthy crops. The detection of plant diseases is of fundamental importance in practical agricultural production. It controls the growth and health of the plant and ensures the regular operation and successful harvest of agricultural plantations. The disease affecting the plants is determined by factors such as the climate. This paper examines an alternative approach to developing a disease detection model supported by leaf classification using deep convolutional networks. Growth in computer vision present a scope to broaden and boost the practice of precision crop protection and expand the market for computer vision applications in precision agriculture, a completely unique form of training and therefore the technique used allows for quick and direct implementation of the system in practice. The database used in this paper consists of 77,000 images of healthy and infected plant leaves. We were able to train a CNN model for classifying plant diseases that is, they are present or not, and then another model was trained with YOLOv7 to detect the disease. The trained classification model achieved an accuracy of 99.5% and the detection model was able to achieve mA <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$P$</tex> , precision, recall of 0.65, 0.59and 0.65 respectively.