Biomedical Engineering Reference
In-Depth Information
Fig. 6.18 A sequence of images recorded during the real time stereo tracking of a person walking
in front of the iCub: six images are shown in temporal order from L1 to L6 (
left camera
) and R1 to
R6 (
right camera
). The walking person is highlighted with a
green blob
using the result of
proposed algorithm
6.8 Conclusions
This chapter deals with the problem of building a reliable architecture to control
movement by relying on sensory data in a humanoid robot where many degrees of
freedom need to be coordinated. We have shown original solutions to vision (using
motion), to kinematics (using robust optimization and a multi-referential trajectory
formulation), and to dynamics (by enabling impedance control from a set of FTSs
and tactile sensors). Although certain aspects of these methods are somewhat
traditional, their specific application and combination is novel. We took particular
care in testing all methods rigorously and comparing them with other methods in
the literature.
Furthermore, the entire implementation of this software is available, following
the iCub policies, as open source (GPL) from the iCub repository. These libraries
and modules, besides running on the iCub, are available to the research community
at large. The algorithms are almost always embedded in static libraries ready to be
picked up by others.
The iCub repository can be found at
http://www.icub.org
and browsed on
Source-Forge (
http://www.sourceforge.net
)
. Several videos of the iCub showing
the methods described in this paper are available on the Internet and in particular at
this site:
http://www.youtube.com/robotcub
.
References
Abend W, Bizzi E, Morasso P (1982) Human arm trajectory formation. Brain 105:331-348
Arbib MA (1981) Perceptual structures and distributed motor control. In: Brooks VB
(ed) Handbook of physiology, vol. II, motor control. American Physiological Society,
Bethesda, MD, pp 1449-1480
Barequet G, Har-Peled S (2001) Efficiently approximating the minimum-volume bounding box of
a point set in three dimensions. J Algorithm 38:91-109
Bjorck A (1996) Numerical methods for least squares problems. Society for Industrial Mathematics,
Philadelphia, PA
Search WWH ::
Custom Search