Kinematic model aided inertial motion tracking of human upper limb

Abstract

A new motion tracking framework has been developed to estimate the position and orientation of human upper limb. This method fuses data from on-board accelerometers and gyroscopes, which are accommodated in a commercially available inertial sensor MT9. Human upper limb motion can be represented by a kinematic chain in which six joint variables are to be considered: three for the shoulder and three for the elbow. Based on measurements of the inertial sensor placed on the wrist, we then obtain the positions of the wrist and elbow. An extended Kalman filter then fuses the data from these sensors in order to reduce errors and noise in measurements. Preliminary results demonstrate the favorable performance of the proposed strategy.