DefinePK

Abstract

Cotton is a vital cash crop globally, and its health and productivity are constantly threatened by various diseases. Early detection and accurate diagnosis of these diseases are crucial for effective crop management and minimizing yield losses. In this study, we propose a cotton leaf disease detection system utilizing object detection techniques. Creating an accurate, automated system for spotting and locating illnesses on cotton leaves is the aim of this study. Due to its real-time processing capabilities, we use cutting-edge object detection algorithms, concentrating on the widely used YOLO (You Only Look Once) paradigm. The model is trained using a sizable dataset of cotton leaf photos that have been annotated and creating an xml file and contain samples that have disease infections (fungal). The proposed approach utilizes the ResNet-101 deep convolutional neural network, which has demonstrated strong performance in various computer vision tasks. The model is pretrained on large-scale image datasets to capture high-level features and then fine-tuned on a custom dataset containing annotated cotton leaf images. The dataset used in this research consists of diverse images of cotton plants captured under various environmental conditions. Each image is manually annotated to mark the bounding boxes around individual cotton leaves. These annotations serve as ground truth data for training and evaluating the object detection model. our proposed model achieved an accuracy of 93 percent.