A six point solution for structure and motion

The paper has two main contributions: The first is a set of methods for computing structure and motion for m ≥ 3 views of 6 points. It is shown that a geometric image error can be minimized over all views by a simple three parameter numerical optimization. Then, that an algebraic image error can be...

Full description

Bibliographic Details
Main Authors: Schaffalitzky, F, Zisserman, A, Hartley, RI, Torr, PHS
Format: Conference item
Language:English
Published: Springer 2003
_version_ 1817931379488325632
author Schaffalitzky, F
Zisserman, A
Hartley, RI
Torr, PHS
author_facet Schaffalitzky, F
Zisserman, A
Hartley, RI
Torr, PHS
author_sort Schaffalitzky, F
collection OXFORD
description The paper has two main contributions: The first is a set of methods for computing structure and motion for m ≥ 3 views of 6 points. It is shown that a geometric image error can be minimized over all views by a simple three parameter numerical optimization. Then, that an algebraic image error can be minimized over all views by computing the solution to a cubic in one variable. Finally, a minor point, is that this “quasi-linear" linear solution enables a more concise algorithm, than any given previously, for the reconstruction of 6 points in 3 views. The second contribution is an m view n ≥ 6 point robust reconstruction algorithm which uses the 6 point method as a search engine. This extends the successful RANSAC based algorithms for 2-views and 3-views to m views. The algorithm can cope with missing data and mismatched data and may be used as an efficient initializer for bundle adjustment. The new algorithms are evaluated on synthetic and real image sequences, and compared to optimal estimation results (bundle adjustment).
first_indexed 2024-12-09T03:21:05Z
format Conference item
id oxford-uuid:314485fb-2e66-4f2d-9feb-948a57552a68
institution University of Oxford
language English
last_indexed 2024-12-09T03:21:05Z
publishDate 2003
publisher Springer
record_format dspace
spelling oxford-uuid:314485fb-2e66-4f2d-9feb-948a57552a682024-11-12T15:47:04ZA six point solution for structure and motionConference itemhttp://purl.org/coar/resource_type/c_5794uuid:314485fb-2e66-4f2d-9feb-948a57552a68EnglishSymplectic ElementsSpringer2003Schaffalitzky, FZisserman, AHartley, RITorr, PHSThe paper has two main contributions: The first is a set of methods for computing structure and motion for m ≥ 3 views of 6 points. It is shown that a geometric image error can be minimized over all views by a simple three parameter numerical optimization. Then, that an algebraic image error can be minimized over all views by computing the solution to a cubic in one variable. Finally, a minor point, is that this “quasi-linear" linear solution enables a more concise algorithm, than any given previously, for the reconstruction of 6 points in 3 views. The second contribution is an m view n ≥ 6 point robust reconstruction algorithm which uses the 6 point method as a search engine. This extends the successful RANSAC based algorithms for 2-views and 3-views to m views. The algorithm can cope with missing data and mismatched data and may be used as an efficient initializer for bundle adjustment. The new algorithms are evaluated on synthetic and real image sequences, and compared to optimal estimation results (bundle adjustment).
spellingShingle Schaffalitzky, F
Zisserman, A
Hartley, RI
Torr, PHS
A six point solution for structure and motion
title A six point solution for structure and motion
title_full A six point solution for structure and motion
title_fullStr A six point solution for structure and motion
title_full_unstemmed A six point solution for structure and motion
title_short A six point solution for structure and motion
title_sort six point solution for structure and motion
work_keys_str_mv AT schaffalitzkyf asixpointsolutionforstructureandmotion
AT zissermana asixpointsolutionforstructureandmotion
AT hartleyri asixpointsolutionforstructureandmotion
AT torrphs asixpointsolutionforstructureandmotion
AT schaffalitzkyf sixpointsolutionforstructureandmotion
AT zissermana sixpointsolutionforstructureandmotion
AT hartleyri sixpointsolutionforstructureandmotion
AT torrphs sixpointsolutionforstructureandmotion