Upper body motion tracking with inertial sensors

We introduce a motion tracking method of arbitrary human upper body motion. Low cost wearable inertial sensors are employed in our approach to track the upper body movement in 3D space and in real time. We first establish a kinematic chain of a human upper body consisting of a trunk and two limbs. Then joint variables are computed for given rotation matrices of body segments which are measured using 6 inertial measurement units. For this purpose, we solve a reduced form of inverse kinematics in the Lie group setting. To compare efficiency and accuracy of our system with the commercial marker based optical tracking system, Hawk Digital Real Time System, the upper body motion experiment was performed using the proposed algorithm.