Comparison of Orientation Filter Algorithms for Realtime Wireless Inertial Posture Tracking

Advances in the miniaturisation of inertial sensors have allowed the design of compact wireless inertial orientation trackers. Such devices require data fusion algorithms to process sensor data into estimated orientations. This paper examines the problem of inertial sensor data fusion and compares two alternative methods for orientation estimation: complementary filtering and Kalman filtering. Experiments are presented to assess the performance and accuracy of the resulting filters. The complementary filter structure is demonstrated to require up to nine times less execution time, while maintaining better accuracy across different movement scenarios, than the Kalman filter structure.