A robust landmark-based system for vehicle location using low-bandwidth vision

https://doi.org/10.1016/S0921-8890(98)00005-0Get rights and content

Abstract

This paper presents novel computer algorithms, a system architecture, and the prototype implementation of a vision-based automatic vehicle location system. The objective of the vehicle location system is to keep track of the vehicle location for a human driver, and perhaps to provide the driver with real-time audio directions to his destination. The techniques developed here are equally applicable to autonomous robot navigation. The prototype system uses odometer readings and a skeleton map to perform dead reckoning, and uses low-bandwidth visual information and neural networks to recognize places for correcting cumulative dead reckoning errors. The visual information is also used to detect turns, for dead reckoning at intersections. The system is self-contained in the sense that it requires no infrastructure outside the vehicle, such as external beacons installed on roadways or satellites used by Global Positioning Systems (GPS). The system maintains a large number of location hypotheses and searches for a large number of landmarks stored in a database in real time. Hence the system is robustly able to recover the vehicle location after being lost for various reasons. The system has been tested, with success, in both day and night time, in all four seasons, and on roads in New York City, a regional highway, and on suburban streets.

References (13)

  • S.P. Engelson

    Active place recognition using image signatures

  • R.E. Gorr et al.

    A vision based system for automatic vehicle location

  • T.R. Hancock et al.

    Automatic vehicle location using sensory maps

  • J. Hong et al.

    Image-based homing

  • J.S. Judd et al.

    Discoidal visual image detector

    Attorney Docket No. 94E7618

    (December 1994)
  • B.J. Kuipers et al.

    A robust qualitative method for spatial learning

There are more references available in the full text version of this article.

Cited by (9)

  • A neural network based obstacle-navigation animat in a virtual environment

    2002, Engineering Applications of Artificial Intelligence
  • Voronoi based place recognition using hidden Markov models

    2002, IFAC Proceedings Volumes (IFAC-PapersOnline)
  • The state of play in machine/environment interactions

    2006, Artificial Intelligence Review
  • Symbolic place recognition in Voronoi-based maps by using hidden Markov models

    2004, Journal of Intelligent and Robotic Systems: Theory and Applications
  • Robot localization using omnidirectional color images

    2001, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
View all citing articles on Scopus
View full text