Kalman filtering for sensor fusion in a human tracking system

Abstract

This chapter presents a human tracking system for developing human-robot interaction tasks. This system is composed by two subsystems: an inertial motion capture system and a Ultra-WideBand (UWB) localization system. Two fusion algorithms are used to combine the measurements of both systems. The first fusion algorithm transforms measurements from the two systems in the same coordinate system by recalculating the transformation matrix each time a new measurement from the UWB system is received. This approach relies heavily on the accuracy of the measurements from the UWB system because the transformation matrix recalculation assumes that the last UWB measurement is completely correct. Thus, errors in UWB measurements are not considered and only the translational errors of the motion capture system are corrected. The second algorithm takes into account UWB errors and overcomes the drawbacks of the first approach by adding a modified Kalman filter.