Classification of body movements based on posturographic data.

Area of Science:

• Biomechanics
• Human postural control
• Sensor technology

Background:

• The human body exhibits continuous sway during standing.
• Subtle changes in sway, reflecting alterations in ground reaction forces and center of pressure (COP), are influenced by movements, sensory input, emotions, and health.
• Understanding these sway patterns is crucial for analyzing human movement and balance.

Purpose of the Study:

• To demonstrate that specific COP parameters and classification techniques can distinguish between distinct body movements during standing.
• To explore the potential of COP analysis for camera-free motion identification.
• To advance the field of human motion analysis through non-invasive methods.

Main Methods:

• Collected force platform data from participants executing 11 distinct postural and gestural movements.
• Extracted 23 displacement- and frequency-based features from COP time series.
• Employed linear and nonlinear classifiers, including linear discriminants, nearest neighbor classifiers, and support vector machines, for movement identification.

Main Results:

• Achieved average classification rates ranging from 67% to 100% on unseen test data.
• Demonstrated the effectiveness of COP feature extraction and classification in differentiating between various body movements.
• Identified that no single method universally outperformed others across all movement types.

Conclusions:

• Selected COP parameters and classification methods can effectively differentiate specific body movements during standing.
• This approach offers promising prospects for camera-free motion identification.
• A combination of movement-specific features and classifiers is recommended for optimal performance in identifying distinct postural and gestural movements.