Affine correspondences between central cameras for rapid relative pose estimation

Iván Eichhardt, D. Chetverikov

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

Abstract

This paper presents a novel algorithm to estimate the relative pose, i.e. the 3D rotation and translation of two cameras, from two affine correspondences (ACs) considering any central camera model. The solver is built on new epipolar constraints describing the relationship of an AC and any central views. We also show that the pinhole case is a specialization of the proposed approach. Benefiting from the low number of required correspondences, robust estimators like LO-RANSAC need fewer samples, and thus terminate earlier than using the five-point method. Tests on publicly available datasets containing pinhole, fisheye and catadioptric camera images confirmed that the method often leads to results superior to the state-of-the-art in terms of geometric accuracy.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
EditorsMartial Hebert, Yair Weiss, Vittorio Ferrari, Cristian Sminchisescu
PublisherSpringer Verlag
Pages488-503
Number of pages16
ISBN (Print)9783030012304
DOIs
Publication statusPublished - Jan 1 2018
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: Sep 8 2018Sep 14 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11210 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th European Conference on Computer Vision, ECCV 2018
CountryGermany
CityMunich
Period9/8/189/14/18

Fingerprint

Pose Estimation
Correspondence
Camera
Cameras
RANSAC
Robust Estimators
Terminate
Specialization
Estimate
Model

Keywords

  • Affine correspondences
  • Central cameras
  • Relative pose

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Eichhardt, I., & Chetverikov, D. (2018). Affine correspondences between central cameras for rapid relative pose estimation. In M. Hebert, Y. Weiss, V. Ferrari, & C. Sminchisescu (Eds.), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings (pp. 488-503). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11210 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-01231-1_30

Affine correspondences between central cameras for rapid relative pose estimation. / Eichhardt, Iván; Chetverikov, D.

Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. ed. / Martial Hebert; Yair Weiss; Vittorio Ferrari; Cristian Sminchisescu. Springer Verlag, 2018. p. 488-503 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11210 LNCS).

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

Eichhardt, I & Chetverikov, D 2018, Affine correspondences between central cameras for rapid relative pose estimation. in M Hebert, Y Weiss, V Ferrari & C Sminchisescu (eds), Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11210 LNCS, Springer Verlag, pp. 488-503, 15th European Conference on Computer Vision, ECCV 2018, Munich, Germany, 9/8/18. https://doi.org/10.1007/978-3-030-01231-1_30
Eichhardt I, Chetverikov D. Affine correspondences between central cameras for rapid relative pose estimation. In Hebert M, Weiss Y, Ferrari V, Sminchisescu C, editors, Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. Springer Verlag. 2018. p. 488-503. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-01231-1_30
Eichhardt, Iván ; Chetverikov, D. / Affine correspondences between central cameras for rapid relative pose estimation. Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings. editor / Martial Hebert ; Yair Weiss ; Vittorio Ferrari ; Cristian Sminchisescu. Springer Verlag, 2018. pp. 488-503 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{630efa3ef0454da085663dab3dce592a,
title = "Affine correspondences between central cameras for rapid relative pose estimation",
abstract = "This paper presents a novel algorithm to estimate the relative pose, i.e. the 3D rotation and translation of two cameras, from two affine correspondences (ACs) considering any central camera model. The solver is built on new epipolar constraints describing the relationship of an AC and any central views. We also show that the pinhole case is a specialization of the proposed approach. Benefiting from the low number of required correspondences, robust estimators like LO-RANSAC need fewer samples, and thus terminate earlier than using the five-point method. Tests on publicly available datasets containing pinhole, fisheye and catadioptric camera images confirmed that the method often leads to results superior to the state-of-the-art in terms of geometric accuracy.",
keywords = "Affine correspondences, Central cameras, Relative pose",
author = "Iv{\'a}n Eichhardt and D. Chetverikov",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-01231-1_30",
language = "English",
isbn = "9783030012304",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "488--503",
editor = "Martial Hebert and Yair Weiss and Vittorio Ferrari and Cristian Sminchisescu",
booktitle = "Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings",

}

TY - GEN

T1 - Affine correspondences between central cameras for rapid relative pose estimation

AU - Eichhardt, Iván

AU - Chetverikov, D.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - This paper presents a novel algorithm to estimate the relative pose, i.e. the 3D rotation and translation of two cameras, from two affine correspondences (ACs) considering any central camera model. The solver is built on new epipolar constraints describing the relationship of an AC and any central views. We also show that the pinhole case is a specialization of the proposed approach. Benefiting from the low number of required correspondences, robust estimators like LO-RANSAC need fewer samples, and thus terminate earlier than using the five-point method. Tests on publicly available datasets containing pinhole, fisheye and catadioptric camera images confirmed that the method often leads to results superior to the state-of-the-art in terms of geometric accuracy.

AB - This paper presents a novel algorithm to estimate the relative pose, i.e. the 3D rotation and translation of two cameras, from two affine correspondences (ACs) considering any central camera model. The solver is built on new epipolar constraints describing the relationship of an AC and any central views. We also show that the pinhole case is a specialization of the proposed approach. Benefiting from the low number of required correspondences, robust estimators like LO-RANSAC need fewer samples, and thus terminate earlier than using the five-point method. Tests on publicly available datasets containing pinhole, fisheye and catadioptric camera images confirmed that the method often leads to results superior to the state-of-the-art in terms of geometric accuracy.

KW - Affine correspondences

KW - Central cameras

KW - Relative pose

UR - http://www.scopus.com/inward/record.url?scp=85055121155&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055121155&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-01231-1_30

DO - 10.1007/978-3-030-01231-1_30

M3 - Conference contribution

AN - SCOPUS:85055121155

SN - 9783030012304

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 488

EP - 503

BT - Computer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings

A2 - Hebert, Martial

A2 - Weiss, Yair

A2 - Ferrari, Vittorio

A2 - Sminchisescu, Cristian

PB - Springer Verlag

ER -