Unscented FastSLAM: A Robust Algorithm for the Simultaneous Localization and Mapping Problem

Rao-Blackwellized particle filter and FastSLAM have become popular tools to solve the simultaneous localization and mapping (SLAM) problem. However, the above techniques have two important potential limitations, which are the derivation of the Jacobian matrices and the linear approximations to the nonlinear functions. Also, one of the major challenges of both Rao-Blackwellized particle filter and FastSLAM is to reduce the number of particles while maintaining the estimation accuracy. This paper proposes a new algorithm based on unscented transformation called Unscented FastSLAM (UFastSLAM) that overcomes important drawbacks of the Rao-Blackwellized particle filter frameworks by directly using nonlinear relations. Experimental results in large scale environments are presented that demonstrate the effectiveness of the UFastSLAM algorithm over the previous approaches.