This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Nieto, J. Articles by Nebot, E. Search for Related Content Social Bookmarking                         What's this?

DenseSLAM: Simultaneous Localization and Dense Mapping

ARC Centre of Excellence for Autonomous Systems (CAS), The University of Sydney, NSW, Australia, j.nieto{at}acfr.usyd.edu.au

Jose Guivant

ARC Centre of Excellence for Autonomous Systems (CAS), The University of Sydney, NSW, Australia

Eduardo Nebot

ARC Centre of Excellence for Autonomous Systems (CAS), The University of Sydney, NSW, Australia

This paper addresses the problem of environment representation for Simultaneous Localization and Mapping (SLAM) algorithms.

One of the main problems of SLAM is how to interpret and synthesize the external sensory information into a representation of the environment that can be used by the mobile robot to operate autonomously. Traditionally, SLAM algorithms have relied on sparse environment representations. However, for autonomous navigation, a more detailed representation of the environment is necessary, and the classic feature-based representation fails to provide a robot with sufficient information. While a dense representation is desirable, it has not been possible for SLAM paradigms.

This paper presents DenseSLAM, an algorithm to obtain and maintain detailed environment representations. The algorithm represents different sensory information in dense multi-layered maps. Each layer can represent different properties of the environment, such as occupancy, traversability, elevation or each layer can describe the same environment property using different representations. Implementations of the algorithm with two different representations for the dense maps are shown.

A rich representation has several potential advantages to assist the navigation process, for example to facilitate data association using multi-dimensional maps. This paper presents two particular applications to improve the localization process; the extraction of complex landmarks from the dense maps and the detection of areas with dynamic objects. The paper also presents an analysis of consistency of the maps obtained with DenseSLAM. The position error in the dense maps is analyzed and a method to select the landmarks in order to minimize these errors is explained.

The algorithm was tested with outdoor experimental data taken with a ground vehicle. The experimental results show that the algorithm can obtain dense environment representations and that the detailed representation can be used to improve the vehicle localization process.

Key Words: simultaneous localization and mapping (SLAM) • EKF-SLAM • autonomous mapping

The International Journal of Robotics Research, Vol. 25, No. 8, 711-744 (2006)
DOI: 10.1177/0278364906067379


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?