Improving 3D visualisation interactivity by using octrees

Authors

  • R. E. Loke Vision Laboratory, Dept. of Electronics and Computer Science, University of Algarve
  • R. Lam Vision Laboratory, Dept. of Electronics and Computer Science, University of Algarve
  • F. Rotaru Vision Laboratory, Dept. of Electronics and Computer Science, University of Algarve
  • J. M. H. du Buf Vision Laboratory, Dept. of Electronics and Computer Science; University of Algarve

Keywords:

Segmentation, Surface extraction, 3D visualisation, Octrees, Interactivity, MRI data

Abstract

This paper describes an interactive method for the visualisation of 3D tomographic data with an emphasis on segmentation and surface construction. In the method multiresolution data representations are employed which have the advantage that results at low spatial resolutions are quickly obtained. This facilitates a fast user interaction, even for huge data volumes, and allows a fast selection of a ROI in the data by avoiding computationally-expensive processing at high spatial resolutions. The disadvantage of the multiresolution processing is that the spatial connectivity of the boundary is partly lost at low resolutions. This is recovered by an accurate refinement technique which improves the connectivity at high resolutions by performing a spatial filtering on the object boundaries.

References

Downloads

Published

2000-01-01

Issue

Section

Special section: workshop on Medical Imaging