Recovery of Epipolar Geometry as a Manifold Fitting Problem

Authors:
Liran Goshen;Ilan Shimshoni;P. Anandan;Daniel Keren
Affiliations:
-;-;-;-
Venue:
ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
Year:
2003

Citing 0
Cited 3

Motion Recovery by Integrating over the Joint Image Manifold

International Journal of Computer Vision
FOE estimation: Can image measurement errors be totally "corrected" by the geometric method?

Pattern Recognition
A Probabilistic Method for Point Matching in the Presence of Noise and Degeneracy

Journal of Mathematical Imaging and Vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

The introduction of the joint image manifold allows to treatthe problem of recovering camera motion and epipolar geometryas the problem of fitting a manifold to the data measuredin a stereo pair. The manifold has a singularity andboundary, therefore care must be taken when fitting it.This paper reviews the notion of joint image manifold,and how previous motion recovery methods can be viewedin its context, and then offers a new fitting method, whichimproves upon previous results, especially when the extentof the data and/or the motion are small.