Department of Computer Science Technische Universitaet Muenchen
Home People Teaching Research Publications Info Sitemap
About | Research | Publications | Announcements | Links

Publications

Back to Publication Index

Radu Bogdan Rusu, Jan Bandouch, Zoltan Csaba Marton, Nico Blodow, and Michael Beetz. Action Recognition in Intelligent Environments using Point Cloud Features Extracted from Silhouette Sequences. In Proceedings of the IEEE 17th International Symposium on Robot and Human Interactive Communication (RO-MAN), Muenchen, Germany, 2008.

Download

[PDF]1.6MB  

Abstract

In this paper we present our work on human action recognition in intelligent environments. We classify actions by looking at a time-sequence of silhouettes extracted from various camera images. By treating time as the third spatial dimension we generate so-called space-time shapes that contain rich information about the actions. We propose a novel approach for recognizing actions, by representing the shapes as 3D point clouds and estimating feature histograms for them. Preliminary results show that our method robustly derives different classes of actions, even in the presence of large variability in the data, coming from different persons at different time intervals.

Image

BibTeX Entry

@InProceedings{Rusu08ROMAN,
  author    = {Radu Bogdan Rusu and Jan Bandouch and Zoltan Csaba Marton and Nico Blodow and Michael Beetz},
  title     = {{Action Recognition in Intelligent Environments using Point Cloud Features Extracted from Silhouette Sequences}},
  booktitle = {Proceedings of the IEEE 17th International Symposium on Robot and Human Interactive Communication (RO-MAN), Muenchen, Germany},
  year      = {2008},
  abstract  = {
               In this paper we present our work on human action recognition in intelligent
               environments. We classify actions by looking at a time-sequence of
               silhouettes extracted from various camera images. By treating time as the
               third spatial dimension we generate so-called space-time shapes that contain
               rich information about the actions. We propose a novel approach for
               recognizing actions, by representing the shapes as 3D point clouds and
               estimating feature histograms for them. Preliminary results show that our
               method robustly derives different classes of actions, even in the presence
               of large variability in the data, coming from different persons at different
               time intervals.
  },
}

Based on bib2html.pl (written by Patrick Riley ) on Thu Sep 18, 2008 11:08:07 UTC


Computer Science IX | Technische Universität München