Development of Computer Vision Based Face Tracking Measurement for Sitting Ergonomics

Sitting causes major health problems. Both bad sitting posture and prolonged sitting time increases the likelihood of back pain. The object of our research was to study if it is possible to measure sitting ergonomics using a webcam video. The webcam placed in front of a subject was tracking the movements of subject’s face, which were used to derive the pose of subject’s head. Different algorithms to detect bad posture based on head pose and movement were developed according to the physiological recommendations. The accuracy of these algorithms were investigated in a real environment testing. Although the results are promising, more research is needed to develop better algorithms for a more accurate and versatile detection of sitting ergonomics.