TY - CHAP
T1 - Identification and decomposition of error in 3D motion capture using inertial and magnetic sensors
AU - Lambrecht, S.
AU - Jonkers, I.
AU - Pons, Jose L
N1 - Funding Information:
This work is part of HYPER project, funded by the Spanish government and coordinated by the bioengineering group of CSIC. The goal of the HYPEr project is to desgn and develop neuro-robotic and neuro-prosthetic interventions for rehabilitation and functional compensation of motor function at upper and lower extremities suffering from stroke or SCI CSD2009-00067.
Publisher Copyright:
© 2013, Springer-Verlag Berlin Heidelberg.
PY - 2013
Y1 - 2013
N2 - In this study the ability of inertial and magnetic sensors (TechMCS, Technaid SL, Madrid, Spain) to accurately track 3D motion is evaluated. An effort is made to decompose the accumulated error present in complex motion capture in order to better understand the strengths, weaknesses and fields of improvement of these sensors. A series of tests were conducted in static, quasi-static and pendulum-dynamic conditions. No significant influence of magnetic field distortion or accelerometer drift was observed. Our data suggests that the drift present after integrating the gyroscope data is a function of motion complexity, with an additional random component. Accurate calibration of gyroscopes and improved drift models are expected to further increase performance of wearable motion capture systems. A maximal discrepancy of almost 4° was observed under ideal testing conditions.
AB - In this study the ability of inertial and magnetic sensors (TechMCS, Technaid SL, Madrid, Spain) to accurately track 3D motion is evaluated. An effort is made to decompose the accumulated error present in complex motion capture in order to better understand the strengths, weaknesses and fields of improvement of these sensors. A series of tests were conducted in static, quasi-static and pendulum-dynamic conditions. No significant influence of magnetic field distortion or accelerometer drift was observed. Our data suggests that the drift present after integrating the gyroscope data is a function of motion complexity, with an additional random component. Accurate calibration of gyroscopes and improved drift models are expected to further increase performance of wearable motion capture systems. A maximal discrepancy of almost 4° was observed under ideal testing conditions.
UR - http://www.scopus.com/inward/record.url?scp=85027266256&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85027266256&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-34546-3_117
DO - 10.1007/978-3-642-34546-3_117
M3 - Chapter
AN - SCOPUS:85027266256
T3 - Biosystems and Biorobotics
SP - 721
EP - 725
BT - Biosystems and Biorobotics
PB - Springer International Publishing
ER -