Feature Based Condensation for Mobile Robot Localization

Patric Jensfelt, David Austin, Olle Wijk and Magnus Andersson

Abstract

Much attention has been given to Monte Carlo methods for mobile robot localization. This has resulted in somewhat of a breakthrough in representing uncertainty for mobile robots. In this paper we use CONDENSATION as a tool for doing feature based global localization in a large and semi-structured environment. This paper presents a comparison of four different feature types: sonar based triangulation points and pairs of triangulation points, as well as lines and doors extracted using a laser scanner. We show experimental results that highlight the information content of the different features, and point to fruitful combinations. Four performance measures are introduced to evaluate the results, convergence speed, accuracy, computation time and how well suited the features are for using in combination with what is called planned sampling, an enhancement over random sampling. From the comparison of the features, some general guidelines are drawn for determining good feature types.

BibTeX Entry:

@InProceedings{Jensfelt00a,
  author = 	 {Patric Jensfelt and David Austin and Olle Wijk and Magnus Andersson},
  title = 	 {Feature Based Condensation for Mobile Robot Localization},
  booktitle = 	 {IEEE Intl. Conf. on Robotics and Automation},
  year =	 2000
}

Download: ps.gz (144k) or pdf (386k)