Adaptive SLAM algorithm with sampling based on state uncertainty

ArticleinElectronics Letters 47(4):284 - 286 · March 2011with24 Reads
DOI: 10.1049/el.2010.3476 · Source: IEEE Xplore
Since the uncertainty of a robot state changes over time, proposed is an adaptive simultaneous localisation and mapping (SLAM) algorithm based on the Kullback-Leibler distance (KLD) sampling and Markov chain Monte Carlo (MCMC) move step. First, it can adaptively determine the number of required particles by calculating the KLD between the posterior distribution approximated by particles and the true posterior distribution at each step. Secondly, it introduces the MCMC move step to increase the particle variety. Both simulation and experimental results demonstrate that the proposed algorithm can obtain more robust and precise results by computing the number of required particles more accurately than previous algorithms.
  • [Show abstract] [Hide abstract] ABSTRACT: This work addresses the problem of performing large scale SLAM(Simultaneous Localisation And Mapping) with satellite stereo imagery for terrain mapping, using a constant time estimation approach. The approach adopts the relative bundle adjustment approach (RBA) and integrates with it a particle-based framework to obtain a constant time probabilistic pose estimation model. The approach further uses a concept of fuzzy landmark-based similarity between poses to make common landmark identification across poses easier, especially when landmarks are sparsely encountered. In order to achieve robustness under varying environmental conditions, we use Speeded Up Robust Features(SURF) for computing spatial and temporal landmark correspondences across time steps. Finally, we use a fast loop closure approach to reduce drifts and obtain global pose estimates. For simulation study, the robot images are cropped from stereo-pair satellite images at different time steps incorporating errors in the robot’s control information. Extensive experimentation has been carried out to study the robot trajectories and the determination of Digital Elevation Model(DEM), with encouraging findings. We have also compared our work with 6D FastSLAM 2.0 (Thrun et al (2005)) as well as Relative SLAM(RSLAM) due to Mei et al(2010).
    Full-text · Article · Feb 2016