A new practical, high-performance mobile robot localization technique is de
scribed that is motivated by the fact that many man-made environments conta
in substantially flat, visually textured surfaces of persistent appearance.
While the tracking of image regions is much studied in computer vision, ap
pearance is still a largely unexploited localization resource in commercial
ly relevant applications We show how prior appearance models can be used to
enable highly repeatable mobile robot guidance that, unlike commercial alt
ernatives, is both infrastructure-free and free-ranging. Very large-scale m
osaics are constructed and used to localize a mobile robot operating in the
modeled environment. Straightforward techniques from vision-based localiza
tion and mosaicking are used to produce a field-relevant AGV guidance syste
m based only on vision and odometry. The feasibility, design, implementatio
n, and precommercial field qualification of such a guidance system are desc
ribed.