3D Research

, 1:4

Using articulated scene models for dynamic 3d scene analysis in vista spaces

  • Niklas Beuter
  • Agnes Swadzba
  • Franz Kummert
  • Sven Wachsmuth
3DR Express

DOI: 10.1007/3DRes.03(2010)04

Cite this article as:
Beuter, N., Swadzba, A., Kummert, F. et al. 3D Res (2010) 1: 4. doi:10.1007/3DRes.03(2010)04

Abstract

In this paper we describe an efficient but detailed new approach to analyze complex dynamic scenes directly in 3D. The arising information is important for mobile robots to solve tasks in the area of household robotics. In our work a mobile robot builds an articulated scene model by observing the environment in the visual field or rather in the so-called vista space. The articulated scene model consists of essential knowledge about the static background, about autonomously moving entities like humans or robots and finally, in contrast to existing approaches, information about articulated parts. These parts describe movable objects like chairs, doors or other tangible entities, which could be moved by an agent. The combination of the static scene, the self-moving entities and the movable objects in one articulated scene model enhances the calculation of each single part. The reconstruction process for parts of the static scene benefits from removal of the dynamic parts and in turn, the moving parts can be extracted more easily through the knowledge about the background. In our experiments we show, that the system delivers simultaneously an accurate static background model, moving persons and movable objects. This information of the articulated scene model enables a mobile robot to detect and keep track of interaction partners, to navigate safely through the environment and finally, to strengthen the interaction with the user through the knowledge about the 3D articulated objects and 3D scene analysis.

Keywords

Vista space Articulated scene model Mobile robot Person tracking 3D background modeling 

Copyright information

© 3D Display Research Center and Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Niklas Beuter
    • 1
  • Agnes Swadzba
    • 1
  • Franz Kummert
    • 1
  • Sven Wachsmuth
    • 1
  1. 1.Bielefeld UniversityBielefeldGermany

Personalised recommendations