Automated Detection of Improper Sitting Postures in Computer Users Based on Motion Capture Sensors

Prolonged computer-related work can be linked to musculoskeletal disorders (MSD) in the upper limbs and improper posture. In this regard, we report on developing resources supporting posture studies based motion capture sensors. These were used create a baseline detector for automated detection of sitting postures, which was next evaluate applicability Hjorth’s parameters—Activity, Mobility Complexity—on specific classification task. Specifically, accelerometer data, computed time-domain parameters, stacked as feature vectors fed binary classifier (kNN, decision tree, linear SVM Gaussian SVM). The experimental evaluation setup involving two different keyboard types (standard ergonomic) validated practical worth proposed method, reported an average accuracy up 98.4%. We deem that research contributes toward creating system monitoring people working computer prolonged periods.