Tuning stereo image matching with stereo video sequence processing

  • Authors:
  • Andrew Speers;Michael Jenkin

  • Affiliations:
  • York University, Toronto, Ontario;York University, Toronto, Ontario

  • Venue:
  • Proceedings of the 2012 Joint International Conference on Human-Centered Computer Environments
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Algorithms for stereo video image processing typicaly assume that the various tasks; calibration, static stereo matching, and egomotion are independent black boxes. In particular, the task of computing disparity estimates is normally performed independently of ongoing egomotion and environmental recovery processes. Can information from these processes be exploited in the notoriously hard problem of disparity field estimation? Here we explore the use of feedback from the environmental model being constructed to the static stereopsis task. A prior estimate of the disparity field is used to seed the stereomatching process within a probabilistic framework. Experimental results on simulated and real data demonstrate the potential of the approach.