JanusVF: Adaptive Fiducial Selection

  • Authors:
  • Malcolm Hutson;Dirk Reiners

  • Affiliations:
  • University of Louisiana at Lafayette,;University of Louisiana at Lafayette,

  • Venue:
  • ISVC '09 Proceedings of the 5th International Symposium on Advances in Visual Computing: Part I
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Critical limitations exist in the currently available commercial tracking technologies for fully-enclosed virtual reality (VR) systems. While several 6DOF solutions can be adapted for fully-enclosed spaces, they still include elements of hardware that can interfere with the users visual experience. JanusVF introduced a tracking solution for fully-enclosed VR displays that achieves comparable performance to available commercial solutions but without artifacts that obscure the users view. In this paper we extend JanusVF by exploring two new methods for selecting the fiducials to be displayed. The first method creates an additional offset grid of fiducials for each original resolution. The algorithm selects fiducials from the grid that yields more visible markers. The second method selects multiple resolutions of markers per wall by intersecting the cameras view frustum with the display surface. The resulting area is populated with fiducials by a quadtree subdivision. Comparison results show that the positional error has been reduced.