Photorealistic Scene Reconstruction by Voxel Coloring

  • Authors:
  • Steven M. Seitz;Charles R. Dyer

  • Affiliations:
  • The Robotics Institute, Carnegie Mellon University, Pittsburgh, PA 15213;Department of Computer Sciences, University of Wisconsin, Madison, WI 53706

  • Venue:
  • International Journal of Computer Vision
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

A novel scene reconstruction technique is presented,different from previous approaches in its ability to copewith large changes in visibility and its modeling of intrinsicscene color and texture information. The methodavoids image correspondence problems by working in adiscretized scene space whose voxels are traversed in afixed visibility ordering. This strategy takes full accountof occlusions and allows the input cameras to be far apartand widely distributed about the environment. The algorithmidentifies a special set of invariant voxels which togetherform a spatial and photometric reconstruction of thescene, fully consistent with the input images. The approachis evaluated with images from both inward-facing and outward-facingcameras.