Full-Surround Image Display Technologies

  • Authors:
  • Hiroo Iwata

  • Affiliations:
  • Institute of Engineering Mechanics and Systems, University of Tsukuba, Japan

  • Venue:
  • International Journal of Computer Vision - Special Issue on Omni-Directional Research in Japan
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents two methods for developing full-surround displays: rear-projection-based “Garnet Vision” and specialized spherical screen named “Ensphered Vision.” The Garnet Vision is a closed polyhedron screen. A viewer stands inside the polyhedron so that image covers full solid angle around the viewer. Optimum configuration of polyhedron is determined by two criteria: pixel efficiency and space efficiency. The criteria maximizes space utility of the display. Through examination of these criteria, rhombic dodecahedron is chosen. A dodecahedron screen with twelve projectors is built, in which a viewer can stand. Ensphered Vision is an image display system for full-surround spherical screen. Sphere is an ideal shape of a screen that covers human visual field. Distance between eyes and screen should be constant while a viewer rotates the head. We use single projector and a convex mirror in order to display seamless image. The optical system employs two mirrors: a plain mirror and a spherical convex mirror. The spherical convex mirror scatters the light from the projector in the spherical screen. The flat mirror provides the viewer to the sweet-spot where he/she can see the image from the center of the sphere. This optical configuration enables seamless wide-angle image in a very limited space.