In this paper we present an integrated approach for efficient online 3D semantic map building of urban environments and the subsequent extraction of qualitative spatial relationships between the different objects in the scene. We split this process into three stages, where we combine a state of the art image segmentation and classification algorithm with an online clustering algorithm to obtain a coherent representation of the environment. Finally, a graph representation is extracted which can then be used for spatial reasoning and human robot interaction. We present first results from data collected by a mobile robot which operates in city areas. © 2012 Springer-Verlag.
CITATION STYLE
Mitsou, N., De Nijs, R., Lenz, D., Frimberger, J., Wollherr, D., Kühnlenz, K., & Tzafestas, C. (2012). Online semantic mapping of urban environments. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7463 LNAI, pp. 54–73). https://doi.org/10.1007/978-3-642-32732-2_4
Mendeley helps you to discover research relevant for your work.