PrimeSense Sensors as a Low-cost Method for Motion Capture on Clinical Tests

Abstract

Virtual Reality (VR) is the computer recreation of simulated environments that create on the user a sense of physical presence. VR provides the advantages of being highly flexible and controllable, allowing experts to generate the optimal conditions for any given test and isolating any desired variables in the course of an experiment. An important feature of VR is that it allows interaction within the virtual world. Motion capture is one of the most popular technologies, because it contributes to creating in the subject the sense of presence. There are several methods to incorporate these techniques into an VR system, with the challenge of them not being too invasive. We propose a method using PrimeSense sensors and several computer vision techniques to build a low-cost system that has proven to be valid in its application as a support therapy for Parkinson's disease (PD) patients.