Laser Stripe Bathymetry using Particle Filter SLAM

Abstract

This paper presents a novel application of BPSLAM from Barkby et al [1], onto underwater laser mapping. BP-SLAM is a featureless bathymetry mapping algorithm that uses Distributed Particle Mapping and a Rao-Backwelized Particle Filter (RBPF) to handle both navigational estimates and seafloor bathymetry estimates. By weighting the particles in the RBPF with the most self-consistent map, the bathymetric maps and the robot navigation are corrected.