 |
Ultra-Wideband Radio Based Human Motion Tracking
Tracking the movement of a human body is of interest in various disciplines.
In the field of medicine, it has been used to assist patients who undergo a stroke rehabilitation process;
in sports science, motion tracking can be used to analyse athletes training and exercise;
in the entertainment industry, the motion of human actors and animals can be recorded to create an avatar animation.
The main goal of this project is to investigate an Ultra-Wideband (UWB) radio based human motion tracking system.
In the considered system setup, both the agents and the anchors are attached on the body.
The agents are low-cost transmit-only nodes which might be unsynchronized.
On the other hand, anchors are relatively complex transceiver nodes which are synchronized to each other.
By exploiting the very large signal bandwidth (from 3.1 to 10.6 GHz), body mounted UWB nodes can determine their relative distance with very high spatio-temporal resolution
and perform motion tracking based on this metric. Moreover, UWB based system meets the key requirements of wearable computing:
low cost, small form factor and low energy consumption.
Combined with the fact that such systems can be implemented as a secondary application of an existing communication centric body area networks makes UWB
an interesting technology for human motion tracking.
Despite such a system might find application in many fields, substantial research is required to achieve its feasibility and accuracy.
Among the major challenges that need to be overcome are multipath propagation, potential non-line of sight conditions and
limited choice of anchor position which have a strong impact on the localization accuracy.
To circumvent these problems, we exploit the constraints on agents’ position that are imposed by the kinematics of the body.
For example, an elbow joint only has limited degree of freedom as compared to its corresponding shoulder; the forearm and the upper arm have a given lengths.
These and similar information can be investigated to improve localization performance.
Publications
-
Constrained Maximum Likelihood Positioning for UWB Based Human Motion Tracking
Z. W. Mekonnen, E. Slottke, H. Luecken, C. Steiner, and A. Wittneben, 2010 International Conference on Indoor Positioning and Indoor Navigation (IPIN), Zurich, Switzerland, Sept. 2010
Localization via Taylor Series Approximation for UWB based Human Motion Tracking
Z. W. Mekonnen and A. Wittneben, 8th Workshop on Positioning, Navigation and Communication (WPNC), Dresden, Germany, Apr. 2011.
|
|
