• Rob-d sensor
• optoelectronic system
• infant
• features extraction

This work presents a new motion tracking method to evaluate spontaneous movements in preterm infants using an RGB depth sensor (Xtion Pro Live Asus).
The assessment of motor repertoire in preterm infants is a diagnostic tool for an early identification of neurodevelopmental disorders, like cerebral palsy. Nowadays the assessment is based on observations and experiences of the clinicians.
It is evident that a motion tracking system that can support clinicians is needed for an objective assessment.
However, available motion tracking systems are marker-based, invasive and need a controlled environment. In this context, the aim of this thesis is the validation of a markerless system compared to the optoelectronic system, the gold standard for motion tracking. In addition, the new system has been adopted in a clinical experimental study.