Range Segmentation Using Visibility Constraints

  • Authors:
  • Leonid Taycher;Trevor Darrell

  • Affiliations:
  • Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, 02139, USA. lodrion@ai.mit.edu;Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, 02139, USA. trevor@ai.mit.edu

  • Venue:
  • International Journal of Computer Vision
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Visibility constraints can aid the segmentation of foreground objects in a scene observed with multiple range imagers. Points may be labeled as foreground if they can be determined to occlude some space in the scene that we expect to be empty. Visibility constraints from a second range view provide evidence of such occlusions. We present an efficient algorithm to estimate foreground points in each range view using explicit epipolar search. In cases where the background pattern is stationary, we show how visibility constraints from other views can generate virtual background values at points with no valid depth in the primary view. We demonstrate the performance of both algorithms for detecting people in indoor office environments with dynamic illumination variation.