Theory and Practice of Projective Rectification

  • Authors:
  • Richard I. Hartley

  • Affiliations:
  • G.E. CRD, Schenectady, NY, 12309. hartley@crd.ge.com

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

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper gives a new method for image rectification,the process of resampling pairs of stereo images taken from widelydiffering viewpoints in order to produce a pair of “matched epipolarprojections”. These are projections in which the epipolar lines runparallel with the x-axis and consequently, disparities between theimages are in the x-direction only. The method is based on anexamination of the fundamental matrix of Longuet-Higgins which describes theepipolar geometry of the image pair. The approach taken is consistentwith that advocated by Faugeras (1992) of avoiding camera calibration. The paper uses methods of projectivegeometry to determine a pair of 2D projective transformations to be applied tothe two images in order to match the epipolar lines. The advantages includethe simplicity of the 2D projective transformation which allows very fastresampling as well as subsequent simplification in the identification ofmatched points and scene reconstruction.