Image segmentation with adaptive sparse grids

Benjamin Peherstorfer; Julius Adorf; Dirk Pflüger; Hans Joachim Bungartz

doi:10.1007/978-3-319-03680-9_17

Image segmentation with adaptive sparse grids

Benjamin Peherstorfer, Julius Adorf, Dirk Pflüger, Hans Joachim Bungartz

Mathematics

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

Abstract

We present a novel adaptive sparse grid method for unsupervised image segmentation. The method is based on spectral clustering. The use of adaptive sparse grids achieves that the dimensions of the involved eigensystem do not depend on the number of pixels. In contrast to classical spectral clustering, our sparse-grid variant is therefore able to segment larger images. We evaluate the method on real-world images from the Berkeley Segmentation Dataset. The results indicate that images with 150,000 pixels can be segmented by solving an eigenvalue system of dimensions 500 x 500 instead of 150,000 x 150,000.

Original language	English (US)
Title of host publication	AI 2013
Subtitle of host publication	Advances in Artificial Intelligence - 26th Australasian Joint Conference, Proceedings
Pages	160-165
Number of pages	6
DOIs	https://doi.org/10.1007/978-3-319-03680-9_17
State	Published - 2013
Event	26th Australasian Joint Conference on Artificial Intelligence, AI 2013 - Dunedin, Netherlands Duration: Dec 1 2013 → Dec 6 2013

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8272 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	26th Australasian Joint Conference on Artificial Intelligence, AI 2013
Country/Territory	Netherlands
City	Dunedin
Period	12/1/13 → 12/6/13

Keywords

Image segmentation
Out-of-sample extension
Sparse grids

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-03680-9_17

Cite this

Peherstorfer, B., Adorf, J., Pflüger, D., & Bungartz, H. J. (2013). Image segmentation with adaptive sparse grids. In AI 2013: Advances in Artificial Intelligence - 26th Australasian Joint Conference, Proceedings (pp. 160-165). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8272 LNAI). https://doi.org/10.1007/978-3-319-03680-9_17

Image segmentation with adaptive sparse grids. / Peherstorfer, Benjamin; Adorf, Julius; Pflüger, Dirk et al.
AI 2013: Advances in Artificial Intelligence - 26th Australasian Joint Conference, Proceedings. 2013. p. 160-165 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8272 LNAI).

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

Peherstorfer, B, Adorf, J, Pflüger, D & Bungartz, HJ 2013, Image segmentation with adaptive sparse grids. in AI 2013: Advances in Artificial Intelligence - 26th Australasian Joint Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8272 LNAI, pp. 160-165, 26th Australasian Joint Conference on Artificial Intelligence, AI 2013, Dunedin, Netherlands, 12/1/13. https://doi.org/10.1007/978-3-319-03680-9_17

Peherstorfer B, Adorf J, Pflüger D, Bungartz HJ. Image segmentation with adaptive sparse grids. In AI 2013: Advances in Artificial Intelligence - 26th Australasian Joint Conference, Proceedings. 2013. p. 160-165. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-03680-9_17

@inproceedings{45d68c9dc40f4192910bf194050153f2,

title = "Image segmentation with adaptive sparse grids",

abstract = "We present a novel adaptive sparse grid method for unsupervised image segmentation. The method is based on spectral clustering. The use of adaptive sparse grids achieves that the dimensions of the involved eigensystem do not depend on the number of pixels. In contrast to classical spectral clustering, our sparse-grid variant is therefore able to segment larger images. We evaluate the method on real-world images from the Berkeley Segmentation Dataset. The results indicate that images with 150,000 pixels can be segmented by solving an eigenvalue system of dimensions 500 x 500 instead of 150,000 x 150,000.",

keywords = "Image segmentation, Out-of-sample extension, Sparse grids",

author = "Benjamin Peherstorfer and Julius Adorf and Dirk Pfl{\"u}ger and Bungartz, {Hans Joachim}",

year = "2013",

doi = "10.1007/978-3-319-03680-9_17",

language = "English (US)",

isbn = "9783319036793",

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

pages = "160--165",

booktitle = "AI 2013",

note = "26th Australasian Joint Conference on Artificial Intelligence, AI 2013 ; Conference date: 01-12-2013 Through 06-12-2013",

}

TY - GEN

T1 - Image segmentation with adaptive sparse grids

AU - Peherstorfer, Benjamin

AU - Adorf, Julius

AU - Pflüger, Dirk

AU - Bungartz, Hans Joachim

PY - 2013

Y1 - 2013

N2 - We present a novel adaptive sparse grid method for unsupervised image segmentation. The method is based on spectral clustering. The use of adaptive sparse grids achieves that the dimensions of the involved eigensystem do not depend on the number of pixels. In contrast to classical spectral clustering, our sparse-grid variant is therefore able to segment larger images. We evaluate the method on real-world images from the Berkeley Segmentation Dataset. The results indicate that images with 150,000 pixels can be segmented by solving an eigenvalue system of dimensions 500 x 500 instead of 150,000 x 150,000.

AB - We present a novel adaptive sparse grid method for unsupervised image segmentation. The method is based on spectral clustering. The use of adaptive sparse grids achieves that the dimensions of the involved eigensystem do not depend on the number of pixels. In contrast to classical spectral clustering, our sparse-grid variant is therefore able to segment larger images. We evaluate the method on real-world images from the Berkeley Segmentation Dataset. The results indicate that images with 150,000 pixels can be segmented by solving an eigenvalue system of dimensions 500 x 500 instead of 150,000 x 150,000.

KW - Image segmentation

KW - Out-of-sample extension

KW - Sparse grids

UR - http://www.scopus.com/inward/record.url?scp=84893749999&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84893749999&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-03680-9_17

DO - 10.1007/978-3-319-03680-9_17

M3 - Conference contribution

AN - SCOPUS:84893749999

SN - 9783319036793

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

SP - 160

EP - 165

BT - AI 2013

T2 - 26th Australasian Joint Conference on Artificial Intelligence, AI 2013

Y2 - 1 December 2013 through 6 December 2013

ER -

Image segmentation with adaptive sparse grids

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this