Publikation

A SLAM Overview from a User's Perspective

Udo Frese, René Wagner, Thomas Röfer

In: KI - Künstliche Intelligenz, German Journal on Artificial Intelligence - Organ des Fachbereiches "Künstliche Intelligenz" der Gesellschaft für Informatik e.V. (KI) 24 3 Seiten 191-198 Springer 9/2010.

Abstrakt

This paper gives a brief overview on the Simultaneous Localization and Mapping (SLAM) problem from the perspective of using SLAM for an application as opposed to the common view in SLAM research papers that focus on investigating SLAM itself. We discuss different ways of using SLAM with increasing difficulty: for creating a map prior to operation, as a black-box localization system, and for providing a growing online map during operation. We also discuss the common variants of SLAM based on 2-D evidence grids, 2-D pose graphs, 2-D features, 3-D visual features, and 3-D pose graphs together with their pros and cons for applications. We point to implementations available on the Internet and give advice on which approach suits which application from our experience.

freseki10.pdf (pdf, 211 KB )

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence