Sub-pixel data fusion and edge-enhanced distance refinement for 2D/3D images

  • Authors:
  • Marvin Lindner;Martin Lambers;Andreas Kolb

  • Affiliations:
  • Computer Graphics Group, Institute for Vision and Graphics, University of Siegen, D-57068 Siegen, Germany.;Computer Graphics Group, Institute for Vision and Graphics, University of Siegen, D-57068 Siegen, Germany.;Computer Graphics Group, Institute for Vision and Graphics, University of Siegen, D-57068 Siegen, Germany

  • Venue:
  • International Journal of Intelligent Systems Technologies and Applications
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

An important field of research in computer vision is the 3Danalysis and reconstruction of objects and scenes. A rather newtechnology in this context is the Photonic Mixer Device (PMD),based on the Time-Of-Flight principle, which measures full-rangedistance information in real-time. Unfortunately, PMD-based deviceshave still limited resolution and provide only IR intensityinformation. This paper describes a fast algorithmic approach tocombine high resolution colour images with PMD distance data,acquired using a binocular camera set-up. The resulting combinedRGBZ-data not only enhances the visual result, but also representsa basis for advanced data processing, e.g. object recognition withsub-pixel accuracy. A simple but efficient method is used to detectgeometric occlusion caused by the binocular set-up, which otherwisewill lead to false colour assignments. Additionally, we introducean enhanced filtering technique used for the edge-enhanced distancerefinement of the geometry provided by the PMD camera. Thetechnique incorporates a proper handling of boundaries and aniterative refinement approach, which can be used to enhance the2D/3D-fusion accuracy.