Improvements for an Appearance-based SLAM-Approach for Large-scale Environments

In continuation of our previous work on visual, appearance-based localization and mapping, we presented in [5] a novel appearance-based, visual SLAM approach. 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 Filter (RBPF) as state estimator to a real-world suitable, online SLAM approach. In our system, each RBPF particle incrementally constructs its own graph-based environment model which is labeled with visual appearance features (extracted from panoramic 360 snapshots of the environment) and the estimated poses of the places where the snapshots were captured. The essential advantage of this appearance-based SLAM approach is its low memory and computing-time requirements. Therefore, the algorithm is able to perform in real-time. In this paper we improve our algorithm to deal with dynamic changes in the environment which is typical in real-world environments. Furthermore, we describe a method to limit the memory consumption of the environment model that is needed for large maps. Finally, we present the results of SLAM experiments in a dynamical and large environment that investigates the stability and localization accuracy of this SLAM technique.