DUAL OF THE FACTORED SOLUTION TO THE SIMULTANEOUS LOCALIZATION AND MAPPING PROBLEM

The SLAM estimation problem has an interesting factorization that decouples the path estimation and the map building problem, using a particle filter to estimate over the possible robot paths. Although there are several successful approaches to this idea, there is a lack of application to indoor feature based mapping. This paper presents a novel factorization, which is the dual of the existing one, that decouples the SLAM into a map estimation and a localization problem, using a particle filter to estimate over maps. We have implemented and tested this formulation, successfully building feature based maps of indoor environments with particle filters for (up to our knowledge) the first time.