Title: Accuracy assessment of a model-based human motion measurement system, a pilot study

Authors: A.J.K. Oudenhuijzen, T. Seitz, H. Bubb

Addresses: TNO Defence & Safety, Kampweg 5, Soesterberg, 3769 ZG, The Netherlands. ' Lehrstuhl fur Ergonomie, Technische Universitat Munchen, Boltzmannstrasse 15, D-85747 Garching, Germany. ' Lehrstuhl fur Ergonomie, Technische Universitat Munchen, Boltzmannstrasse 15, D-85747 Garching, Germany

Abstract: The goal of this pilot study was to assess the accuracy of a model-based human motion measuring system. We compared motion capture data from a model-based system with those from a marker-based system. A simple reach task involving the left shoulder, upper arm and lower arm was used for the comparison in two subjects. The results revealed various errors in the model-based system. The modelled movements for the upper arm and lower arm laid within the standard error for most of the motion sequences of the marker-based system. The modelled end positions deviated from the actual motion sequence in Y and Z direction with 17 and 23%, respectively. The shoulder movement algorithm, in the model-based system, could not be used. Hence, the shoulder was assumed not to move in the marker-based system. However, despite these errors, there are advantages in using the model-based system. Preparation time for subjects decreases since marker placement is not needed in the model-based system. Hence, model-based motion capturing shows potential to enhance human modelling towards dynamics.

Keywords: anthropometry; accuracy; human modelling; human factors; ergonomics; joint locations; joint angles; motion analysis; motion capture; simulation; reach tasks; motion sequences.

DOI: 10.1504/IJHFMS.2006.011681

International Journal of Human Factors Modelling and Simulation, 2006 Vol.1 No.1, pp.40 - 51

Published online: 14 Dec 2006 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article