A study on estimation of seating face geometry using belt-shaped inertial sensors system

Abstract

This paper proposes the method for estimating seating face geometry in dynamic condition using belt-shaped inertial sensors system. This method estimates the posture information in relative coordinate system, and the method compensates the accumulated error by using the adaptive Kalman filter. The sensor fusion estimates the Roll and Pitch angles using gyro sensor and acceleration sensor outputs, and the effect of drift error is reduced. The shape is estimated by applying forward kinematics analysis and the posture information calculated by the sensor fusion. The correction is performed using an adaptive Kalman filter. We conducted the measurement experiments using belt-shaped inertial sensors system and 3D motion capture system in dynamic condition. The results indicated that this method can estimate the face geometry in static and dynamic conditions. This system can be used to represent the face geometry in dynamic condition.