Vision and sonar sensor fusion for mobile robot localization in aliased environments

Monte Carlo localization (MCL) is a common method for self-localization of a mobile robot under the assumption that a map of the environment is available. Original implementations used range sensors like laser scanners and sonar sensors. Recently, localization approaches using vision sensors have been developed with good results. In this paper we compare vision-based with sonar-based MCL approaches in terms of localization accuracy. In particular, we show how in an environment with high perceptual aliasing like our department both approaches bear certain weaknesses while by combining vision and sonar sensors the respective localization errors decrease and overall accuracy is improved