Mobile gait tracking using MS Kinect and robotic platform

Abstract

Computer vision is a fast developing field of science. Activity monitoring and analysis belongs to this topic. Our application is focused on gait monitoring and analysis. The novelty of this attitude is in the combination of the MS Kinect v2 sensor for capturing the gait pattern and mobility of such sensor thanks to placing this device on a robotic platform. Crucial part of this project is the hardware solution and autonomous function of this recording set. The designed platform is a six-wheeled robotic platform controlled by a micro controller using the depth information from the Kinect sensor and the distance driven by robot provided by encoders attached to every motor. The following research is devoted to the proposal of methods for a precise and as narrow as possible control of the platform and post processing of records for a features selection and a classification of physical activities and early diagnostics of gait disorders, primarily the Parkinson’s disease.