YOLOv8 Utilization in Occupational Health and Safety
This research contributes to bridging the gap in links between computer vision technology and concepts of workplace safety by showcasing the effectiveness of YOLOv8 in addressing occupational safety concerns, particularly in the construction industry.
Abstract
The field of occupational health and safety focuses on minimizing losses by preserving and protecting human and physical assets in workplaces. This discipline involves monitoring workplaces and advising management on effective strategies to prevent and reduce losses. Gaining prominence through its remarkable precision and speed, the YOLO ("You-Only-Look-Once") framework assists the progress of dependable and quick object detection in images. Integrated into security systems, live tracking of video feeds for quick identification of doubtful activities, face mask conformity and social distancing are made possible by YOLO. Moreover, YOLO models find application in surface inspection, enhancing quality control in manufacturing. This study investigates the potential of lightweight nano and small YOLOv8 architectures to detect OSHA violations on construction sites, utilizing a publicly available dataset. The results are promising, with the best model achieving a 92% precision rate. YOLOv8 demonstrates potential in enhancing safety measures by detecting safety equipment such as helmets, ensuring worker safety. Despite these possibilities, links between computer vision technology and concepts of workplace safety remain unexplored, with a limited research focus. This research contributes to bridging this gap by showcasing the effectiveness of YOLOv8 in addressing occupational safety concerns, particularly in the construction industry.