Presentation
Preliminary Evaluation of a Smartphone-Based Markerless Motion Capture System for Joint Kinematics Measurement During Symmetric and Asymmetric Lifting Tasks
SessionOE3: Tools and Analysis
DescriptionThis study investigated the feasibility of non-invasive biomechanical assessment of occupational lifting tasks using a deep learning-based markerless motion capture system processing smartphone videos. Seven participants performed standardized symmetric and asymmetric lifting tasks while joint kinematics were collected by both the markerless and optical marker-based motion capture (gold standard) systems. The results showed that the markerless motion capture system provided reasonably accurate joint angles for both symmetric (the average root-mean-squared errors (RMSE): 6.5 ͦ and 8.2 ͦ in the lower and upper extremities, respectively) and asymmetric (the average RMSE: 7.2 ͦ and 9.04 ͦ in the lower and upper extremities, respectively) lifting tasks as compared to the gold standard measures. Given its minimal hardware and software requirements, markerless motion capturing using smartphones can be easily accessible to practitioners and researchers who seek to non-invasively measure biomechanical load associated with occupational lifting tasks in a field setting.
Event Type
Lecture
TimeTuesday, September 10th4:15pm - 4:35pm MST
LocationFLW Salon H
Occupational Ergonomics