Publication Details

SLAM++-A Highly Efficient and Temporally Scalable Incremental SLAM Framework

ILA, V.; POLOK, L.; ŠOLONY, M.; SVOBODA, P. SLAM++-A Highly Efficient and Temporally Scalable Incremental SLAM Framework. INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH, 2017, vol. 2017, no. 1, p. 210-230. ISSN: 1741-3176.
Czech title
SLAM++-Vysoce efektivní a temporálně škálující SLAM framework
Type
journal article
Language
English
Authors
Ila Viorela Simona, Ph.D.
Polok Lukáš, Ing., Ph.D.
Šolony Marek, Ing., Ph.D. (DCGM)
Svoboda Pavel, Ing., Ph.D.
URL
Keywords

nonlinear least squares, incremental covariance recovery, long-term SLAM, loop closure, compact state representation

Abstract

The most common way to deal with the uncertainty present in noisy sensorial perception and action is to model the problem with a probabilistic framework. Maximum likelihood estimation (MLE) is a well-known estimation method used in many robotic and computer vision applications. Under Gaussian assumption, the MLE converts to a nonlinear least squares (NLS) problem. Efficient solutions to NLS exist and they are based on iteratively solving sparse linear systems until convergence. In general, the existing solutions provide only an estimation of the mean state vector, the resulting covariance being computationally too expensive to recover. Nevertheless, in many simultaneous localisation and mapping (SLAM) applications, knowing only the mean vector is not enough. Data association, obtaining reduced state representations, active decisions and next best view are only a few of the applications that require fast state covariance recovery. Furthermore, computer vision and robotic applications are in general performed online. In this case, the state is updated and recomputed every step and its size is continuously growing, therefore, the estimation process may become highly computationally demanding. This paper introduces a general framework for incremental MLE called SLAM++, which fully benefits from the incremental nature of the online applications, and provides efficient estimation of both the mean and the covariance of the estimate. Based on that, we propose a strategy for maintaining a sparse and scalable state representation for large scale mapping. SLAM++ differs from existing implementations by performing all the matrix operations by blocks. This led to extremely fast matrix manipulation and arithmetic operations used in NLS. Even though this paper tests SLAM++ efficiency on SLAM problems, its applicability remains general.

Published
2017
Pages
210–230
Journal
INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH, vol. 2017, no. 1, ISSN 1741-3176
Book
Online First
DOI
UT WoS
000399558300006
EID Scopus
BibTeX
@article{BUT130989,
  author="Viorela Simona {Ila} and Lukáš {Polok} and Marek {Šolony} and Pavel {Svoboda}",
  title="SLAM++-A Highly Efficient and Temporally Scalable Incremental SLAM Framework",
  journal="INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH",
  year="2017",
  volume="2017",
  number="1",
  pages="210--230",
  doi="10.1177/0278364917691110",
  issn="1741-3176",
  url="https://doi.org/10.1177/0278364917691110"
}
Back to top