Globally optimal inlier set maximization with unknown rotation and focal length

Jean Charles Bazin; Yongduek Seo; Richard Hartley; Marc Pollefeys

doi:10.1007/978-3-319-10605-2_52

Globally optimal inlier set maximization with unknown rotation and focal length

Jean Charles Bazin, Yongduek Seo, Richard Hartley, Marc Pollefeys

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

23 Citations (Scopus)

Abstract

Identifying inliers and outliers among data is a fundamental problem for model estimation. This paper considers models composed of rotation and focal length, which typically occurs in the context of panoramic imaging. An efficient approach consists in computing the underlying model such that the number of inliers is maximized. The most popular tool for inlier set maximization must be RANSAC and its numerous variants. While they can provide interesting results, they are not guaranteed to return the globally optimal solution, i.e. the model leading to the highest number of inliers. We propose a novel globally optimal approach based on branch-and-bound. It computes the rotation and the focal length maximizing the number of inlier correspondences and considers the reprojection error in the image space. Our approach has been successfully applied on synthesized data and real images.

Original language	English
Title of host publication	Computer Vision, ECCV 2014 - 13th European Conference, Proceedings
Publisher	Springer Verlag
Pages	803-817
Number of pages	15
Edition	PART 2
ISBN (Print)	9783319106045
DOIs	https://doi.org/10.1007/978-3-319-10605-2_52
Publication status	Published - 2014
Event	13th European Conference on Computer Vision, ECCV 2014 - Zurich, Switzerland Duration: 6 Sept 2014 → 12 Sept 2014

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	8690 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	13th European Conference on Computer Vision, ECCV 2014
Country/Territory	Switzerland
City	Zurich
Period	6/09/14 → 12/09/14

Access to Document

10.1007/978-3-319-10605-2_52

Cite this

Bazin, J. C., Seo, Y., Hartley, R., & Pollefeys, M. (2014). Globally optimal inlier set maximization with unknown rotation and focal length. In Computer Vision, ECCV 2014 - 13th European Conference, Proceedings (PART 2 ed., pp. 803-817). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8690 LNCS, No. PART 2). Springer Verlag. https://doi.org/10.1007/978-3-319-10605-2_52

Bazin, Jean Charles ; Seo, Yongduek ; Hartley, Richard et al. / Globally optimal inlier set maximization with unknown rotation and focal length. Computer Vision, ECCV 2014 - 13th European Conference, Proceedings. PART 2. ed. Springer Verlag, 2014. pp. 803-817 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{c70e5457186543b8b5bd1c41f2ee8440,

title = "Globally optimal inlier set maximization with unknown rotation and focal length",

abstract = "Identifying inliers and outliers among data is a fundamental problem for model estimation. This paper considers models composed of rotation and focal length, which typically occurs in the context of panoramic imaging. An efficient approach consists in computing the underlying model such that the number of inliers is maximized. The most popular tool for inlier set maximization must be RANSAC and its numerous variants. While they can provide interesting results, they are not guaranteed to return the globally optimal solution, i.e. the model leading to the highest number of inliers. We propose a novel globally optimal approach based on branch-and-bound. It computes the rotation and the focal length maximizing the number of inlier correspondences and considers the reprojection error in the image space. Our approach has been successfully applied on synthesized data and real images.",

keywords = "Consensus set maximization, RANSAC, branch-and-bound, inlier detection",

author = "Bazin, {Jean Charles} and Yongduek Seo and Richard Hartley and Marc Pollefeys",

year = "2014",

doi = "10.1007/978-3-319-10605-2_52",

language = "English",

isbn = "9783319106045",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

number = "PART 2",

pages = "803--817",

booktitle = "Computer Vision, ECCV 2014 - 13th European Conference, Proceedings",

address = "Germany",

edition = "PART 2",

note = "13th European Conference on Computer Vision, ECCV 2014 ; Conference date: 06-09-2014 Through 12-09-2014",

}

Bazin, JC, Seo, Y, Hartley, R & Pollefeys, M 2014, Globally optimal inlier set maximization with unknown rotation and focal length. in Computer Vision, ECCV 2014 - 13th European Conference, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 8690 LNCS, Springer Verlag, pp. 803-817, 13th European Conference on Computer Vision, ECCV 2014, Zurich, Switzerland, 6/09/14. https://doi.org/10.1007/978-3-319-10605-2_52

Globally optimal inlier set maximization with unknown rotation and focal length. / Bazin, Jean Charles; Seo, Yongduek; Hartley, Richard et al.
Computer Vision, ECCV 2014 - 13th European Conference, Proceedings. PART 2. ed. Springer Verlag, 2014. p. 803-817 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8690 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Globally optimal inlier set maximization with unknown rotation and focal length

AU - Bazin, Jean Charles

AU - Seo, Yongduek

AU - Hartley, Richard

AU - Pollefeys, Marc

PY - 2014

Y1 - 2014

N2 - Identifying inliers and outliers among data is a fundamental problem for model estimation. This paper considers models composed of rotation and focal length, which typically occurs in the context of panoramic imaging. An efficient approach consists in computing the underlying model such that the number of inliers is maximized. The most popular tool for inlier set maximization must be RANSAC and its numerous variants. While they can provide interesting results, they are not guaranteed to return the globally optimal solution, i.e. the model leading to the highest number of inliers. We propose a novel globally optimal approach based on branch-and-bound. It computes the rotation and the focal length maximizing the number of inlier correspondences and considers the reprojection error in the image space. Our approach has been successfully applied on synthesized data and real images.

AB - Identifying inliers and outliers among data is a fundamental problem for model estimation. This paper considers models composed of rotation and focal length, which typically occurs in the context of panoramic imaging. An efficient approach consists in computing the underlying model such that the number of inliers is maximized. The most popular tool for inlier set maximization must be RANSAC and its numerous variants. While they can provide interesting results, they are not guaranteed to return the globally optimal solution, i.e. the model leading to the highest number of inliers. We propose a novel globally optimal approach based on branch-and-bound. It computes the rotation and the focal length maximizing the number of inlier correspondences and considers the reprojection error in the image space. Our approach has been successfully applied on synthesized data and real images.

KW - Consensus set maximization

KW - RANSAC

KW - branch-and-bound

KW - inlier detection

UR - http://www.scopus.com/inward/record.url?scp=84906508192&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-10605-2_52

DO - 10.1007/978-3-319-10605-2_52

M3 - Conference contribution

SN - 9783319106045

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

SP - 803

EP - 817

BT - Computer Vision, ECCV 2014 - 13th European Conference, Proceedings

PB - Springer Verlag

T2 - 13th European Conference on Computer Vision, ECCV 2014

Y2 - 6 September 2014 through 12 September 2014

ER -

Bazin JC, Seo Y, Hartley R, Pollefeys M. Globally optimal inlier set maximization with unknown rotation and focal length. In Computer Vision, ECCV 2014 - 13th European Conference, Proceedings. PART 2 ed. Springer Verlag. 2014. p. 803-817. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-319-10605-2_52

Globally optimal inlier set maximization with unknown rotation and focal length

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this