This paper concerns simultaneous localization and mapping (SLAM) of large areas. In SLAM the map creation is based on identified landmarks in the environment. When mapping large areas a vast number of landmarks have to be treated, which usually is very time consuming. A common way to reduce the computational complexity is to divide the visited area into submaps, each with a limited number of landmarks. This paper presents a novel method for merging conditionally independent submaps (generated using e.g. EKF-SLAM) by the use of smoothing. By this approach it is possible to build large maps in close to linear time. The approach is demonstrated in two indoor scenarios, where data was collected with a trolley-mounted stereo vision camera. © 2011 Springer-Verlag.
CITATION STYLE
Karlsson, A., Bjärkefur, J., Rydell, J., & Grönwall, C. (2011). Smoothing-based submap merging in large area SLAM. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 6688 LNCS, pp. 134–145). https://doi.org/10.1007/978-3-642-21227-7_13
Mendeley helps you to discover research relevant for your work.