Vision-based measuring system for rider's pose estimation during motorcycle riding

Inertial characteristics of the human body are comparable with the vehicle ones in motorbike riding: the study of a rider's dynamic is a crucial step in system modeling. An innovative vision based system able to measure the six degrees of freedom of the driver with respect to the vehicle is proposed here: the core of the proposed approach is an image acquisition and processing technique capable of reconstructing the position and orientation of a target fixed on the rider's back. The technique is firstly validated in laboratory tests comparing measured and imposed target motion laws and successively tested in a real case scenario during track tests with amateur and professional drivers. The presented results show the capability of the technique to correctly describe the driver's dynamic, his interaction with the vehicle as well as the possibility to use the new measuring technique in the comparison of different driving styles.

► Use of monocular pose estimation for motorcycle rider motion measurement.
► Uncertainty evaluation of the proposed vision based measuring system.
► Applicability of the method to driver–vehicle dynamic interaction study.
► Applicability of the method to different drivers' riding styles study.