Balanced recovery of 3D Structure and Camera motion from Uncalibrated image sequences

Bogdan Georgescu, Peter Meer

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Scopus citations

Abstract

Metric reconstruction of a scene viewed by an uncalibrated camera undergoing an unknown motion is a fundamental task in computer vision. To obtain accurate results all the methods rely on bundle adjustment, a nonlinear optimization technique which minimizes the reprojection error over the structural and camera parameters. Bundle adjustment is optimal for normally distributed measurement noise, however, its performance depends on the starting point. The initial solution is usually obtained by solving a linearized constraint through a total least squares procedure, which yields a biased estimate. We present a more balanced approach where in main computational modules of an uncalibrated reconstruction system, the initial solution is obtained from a statistically justified estimator which assures its unbiasedness. Since the quality of the new initial solution is already comparable with that of the result of bundle adjustment, the burden on the latter is drastically reduced while its reliability is significantly increased. The performance of our system was assessed for both synthetic data and standard image sequences.

Original languageEnglish (US)
Title of host publicationComputer Vision - 7th European Conference on Computer Vision, ECCV 2002, Proceedings
EditorsAnders Heyden, Gunnar Sparr, Mads Nielsen, Peter Johansen
PublisherSpringer Verlag
Pages294-308
Number of pages15
ISBN (Print)9783540437444
DOIs
StatePublished - Jan 1 2002
Event7th European Conference on Computer Vision, ECCV 2002 - Copenhagen, Denmark
Duration: May 28 2002May 31 2002

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2351
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other7th European Conference on Computer Vision, ECCV 2002
CountryDenmark
CityCopenhagen
Period5/28/025/31/02

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Balanced recovery of 3D Structure and Camera motion from Uncalibrated image sequences'. Together they form a unique fingerprint.

  • Cite this

    Georgescu, B., & Meer, P. (2002). Balanced recovery of 3D Structure and Camera motion from Uncalibrated image sequences. In A. Heyden, G. Sparr, M. Nielsen, & P. Johansen (Eds.), Computer Vision - 7th European Conference on Computer Vision, ECCV 2002, Proceedings (pp. 294-308). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2351). Springer Verlag. https://doi.org/10.1007/3-540-47967-8_20