D. Fox, W. Burgard and S. Thrun
Markov Localization for Mobile
Robots in Dynamic Environments
Journal of Artificial Intelligence Research
Abstract
Localization, that is the estimation of a robot'slocation from sensor data,
is a fundamental problem in mobilerobotics. This papers presents a version
of Markov localization whichprovides accurate position estimates and which
is tailored towardsdynamic environments. The key idea of Markov localization
is tomaintain a probability density over the space of all locations of arobot
in its environment. Our approach represents this spacemetrically, using a
fine-grained grid to approximate densities. It isable to globally localize
the robot from scratch and to recover fromlocalization failures. It is robust
to approximate models of theenvironment (such as occupancy grid maps) and
noisy sensors (such asultrasound sensors). Our approach also includes afiltering
techniquewhich allows a mobile robot to reliably estimate its positioneven
indensely populated environments in which crowds of people block therobot's
sensors for extended periods of time. The method describedhere hasbeen implemented
and tested in several real-worldapplications of mobile robots,including the
deployments of two mobilerobots as interactive museum tour-guides.
Download
Full paper
[.ps.gz](2726 kb, 37 pages)
Bibtex
@Article{Fox99Mar,
AUTHOR = {Fox, D. and Burgard, W. and Thrun, S.},
TITLE = {Markov Localization for Mobile Robots in Dynamic
Environments},
JOURNAL = {Journalof Artificial Intelligence Research},
VOLUME = {11},
YEAR = {1999}
}