Publication type: |
Article |
Author: |
U. Frese, M. Hörmann, B. Bäuml, G. Hirzinger |
Title: |
Globally Consistent Visual Localization without a-priori Map (German) |
Volume: |
3 |
Page(s): |
273 – 280 |
Journal: |
automatisierungstechnik |
Year published: |
2000 |
Abstract: |
We present a method that can be used as one module of a mobile robots visual localization system for estimating the robots position without using an a-priori map. It builds a network of uncertain spatial relations using odometry and angular measurements between the robot and certain landmarks, which have to be provided by a visual landmark recognition module that is not covered by this article. The network is modelled as a graph, where each edge represents an uncertain spatial relation and, the uncertainty being denoted set based using ellipsoids and hyperboloids. This method solves the consistency problem of large scale autonomous mapping. |
PDF Version: |
http://www.informatik.uni-bremen.de/agebv/downloads/published/at600frese.pdf |
Keywords: |
Simultaneous Localization and Mapping SLAM mopbile robot navigation covariance intersection set based uncertainty |
Note / Comment: |
(german) |
Status: |
Reviewed |
Last updated: |
28. 10. 2008 |