An Improved Sensor Model on Appearance Based SLAM

Abstract. In our previous work on visual, appearance-based localiza-tion and mapping, we presented in [14] a novel SLAM approach to build visually labeled topological maps. The essential contribution of this work was an adaptive sensor model, which is estimated online, and a graph matching scheme to evaluate the likelihood of a given topological map. Both methods enable the combination of an appearance-based, visual localization and mapping concept with a Rao-Blackwellized Particle Fil-ter (RBPF) as state estimator to a real-world suitable, online SLAM approach. In this paper we improve our algorithm by using a novel prob-ability driven approximation of the local similarity function (the sensor model) to deal with dynamic changes of the appearance in the operation area. 2