A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Computer Vision Techniques for Worker Motion Analysis to Reduce Musculoskeletal Disorders in Construction
Worker health is a serious issue in construction. Injuries and illnesses result in days away from work and incur tremendous costs for construction organizations. Musculoskeletal disorders, in particular, constitute a major category of worker injury. The repetitive movements, awkward postures, and forceful exertions involved in trade work are leading causes of this type of injury. To reduce the number of these injuries, worker activities must be tracked and analyzed. Traditional methods to measure work activities rely upon manual on-site observations which are time-consuming and inefficient. To address these limitations, computer vision techniques for worker motion analysis are proposed to automatically identify non-ergonomic postures and movements without on-site work interruption. Specifically, we intend to acquire 2D skeleton extracting joints from image sequences and, while obtaining 3D coordinates for each joint, reconstruct 3D human skeletons for each frame; these then can be used for diverse ergonomic analyses (e.g., joint angle comparisons with the suggested ergonomic guidelines for trades). In this paper, we therefore discuss how 3D skeleton video images can be reconstructed with two 2D skeleton images recorded from two network surveillance cameras. The results demonstrate that the obtained 3D skeleton video with coordinates of joints have enough detail to be used for motion analysis and have great potential to identify non-ergonomic postures and movements. This information can be used to reduce musculoskeletal disorders in the construction industry.
Computer Vision Techniques for Worker Motion Analysis to Reduce Musculoskeletal Disorders in Construction
Worker health is a serious issue in construction. Injuries and illnesses result in days away from work and incur tremendous costs for construction organizations. Musculoskeletal disorders, in particular, constitute a major category of worker injury. The repetitive movements, awkward postures, and forceful exertions involved in trade work are leading causes of this type of injury. To reduce the number of these injuries, worker activities must be tracked and analyzed. Traditional methods to measure work activities rely upon manual on-site observations which are time-consuming and inefficient. To address these limitations, computer vision techniques for worker motion analysis are proposed to automatically identify non-ergonomic postures and movements without on-site work interruption. Specifically, we intend to acquire 2D skeleton extracting joints from image sequences and, while obtaining 3D coordinates for each joint, reconstruct 3D human skeletons for each frame; these then can be used for diverse ergonomic analyses (e.g., joint angle comparisons with the suggested ergonomic guidelines for trades). In this paper, we therefore discuss how 3D skeleton video images can be reconstructed with two 2D skeleton images recorded from two network surveillance cameras. The results demonstrate that the obtained 3D skeleton video with coordinates of joints have enough detail to be used for motion analysis and have great potential to identify non-ergonomic postures and movements. This information can be used to reduce musculoskeletal disorders in the construction industry.
Computer Vision Techniques for Worker Motion Analysis to Reduce Musculoskeletal Disorders in Construction
Li, Chunxia (author) / Lee, SangHyun (author)
International Workshop on Computing in Civil Engineering 2011 ; 2011 ; Miami, Florida, United States
Computing in Civil Engineering (2011) ; 380-387
2011-06-16
Conference paper
Electronic Resource
English
| British Library Conference Proceedings | 2011
Identifying Modular Construction Worker Tasks Using Computer Vision
| British Library Conference Proceedings | 2021