Self-Calibration and Metric Reconstruction Inspite of Varying and Unknown Intrinsic Camera Parameters

  • Authors:
  • Marc Pollefeys;Reinhard Koch;Luc Van Gool

  • Affiliations:
  • ESAT-PSI, K.U. Leuven, Kardinaal Mercierlaan 94, B-3001 Heverlee, Belgium. Marc.Pollefeys@esat.kuleuven.ac.be;ESAT-PSI, K.U. Leuven, Kardinaal Mercierlaan 94, B-3001 Heverlee, Belgium. Reinhard.Koch@esat.kuleuven.ac.be;ESAT-PSI, K.U. Leuven, Kardinaal Mercierlaan 94, B-3001 Heverlee, Belgium. Luc.VanGool@esat.kuleuven.ac.be

  • Venue:
  • International Journal of Computer Vision - 1998 Marr Prize
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper the theoretical and practical feasibility ofself-calibration in the presence of varying intrinsic camera parameters isunder investigation. The paper‘s main contribution is to propose aself-calibration method which efficiently deals with all kinds of constraintson the intrinsic camera parameters. Within this framework a practical methodis proposed which can retrieve metric reconstruction from image sequencesobtained with uncalibrated zooming/focusing cameras. The feasibility of theapproach is illustrated on real and synthetic examples. Besides this atheoretical proof is given which shows that the absence of skew in the imageplane is sufficient to allow for self-calibration. A counting argument isdeveloped which—depending on the set of constraints—gives the minimum sequence length for self-calibration and a method to detect critical motion sequences is proposed.