Development of an Intelligent CCTV Algorithm for Preventing Musculoskeletal Disorders in Workers During Rebar Operations

Musculoskeletal disorders (MSDs) frequently occur in construction due to repetitive movements and improper postures, yet conventional safety monitoring systems lack real-time analysis capabilities. This study proposes a YOLO-based intelligent CCTV algorithm to monitor worker postures and detect hazardous positions. The system employs YOLOv3 and YOLOv8 for worker detection and integrates Mediapipe Pose Estimation to analyze knee angles. Experimental results show that YOLOv8 outperforms YOLOv3 in detection accuracy (97.96% indoors, 86.25% outdoors) and knee angle recognition (69.34% indoors, 92.40% outdoors). While outdoor lighting conditions reduced helmet and pattern recognition efficiency, the proposed ErgoPose Safety Detection (ESPD) algorithm effectively contributes to MSD prevention. Future improvements will focus on multi-worker recognition and integration with wearable sensors for broader applications across various industries.