Manual lifting task risk evaluation using computer vision system

Low back injury has been a significant problem in industry. Studies indicate that inappropriate design of manual material handling (mmh) task is one of the major factors causing low back pain. Here, an automated system of evaluating the possible risk of low back injury in mmh task is developed. The system applies computer vision technique to identify the working posture, then incorporating biomechanical model and anthropometric data to calculate the low back compression force. By comparing with the specified standard limits, the system can indicate the risk level of the task. Some recommendations will then be provided. The system has been validated by comparing with the existing 2-D software, and it is found to be valid and effecient.