From ramp discontinuities to segmentation tree

Emre Akbas, Narendra Ahuja

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

Abstract

This paper presents a new algorithm for low-level multiscale segmentation of images. The algorithm is designed to detect image regions regardless of their shapes, sizes, and levels of interior homogeneity, by doing a multiscale analysis without assuming any prior models of region geometry. As in previous work, a region is modeled as a homogeneous set of connected pixels surrounded by ramp discontinuities. A new transform, called the ramp transform, is described, which is used to detect ramp discontinuities and seeds for all regions in an image. Region seeds are grown towards the ramp discontinuity areas by utilizing a relaxation labeling procedure. Segmentation is achieved by analyzing the output of this procedure at multiple photometric scales. Finally, all detected regions are organized into a tree data structure based on their recursive containment relations. Experiments on real and synthetic images verify the desired properties of the proposed algorithm.

Original languageEnglish (US)
Title of host publicationComputer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers
Pages123-134
Number of pages12
EditionPART 1
DOIs
StatePublished - Dec 29 2010
Event9th Asian Conference on Computer Vision, ACCV 2009 - Xi'an, China
Duration: Sep 23 2009Sep 27 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume5994 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other9th Asian Conference on Computer Vision, ACCV 2009
CountryChina
CityXi'an
Period9/23/099/27/09

Fingerprint

Discontinuity
Segmentation
Seed
Labeling
Data structures
Pixels
Transform
Multiscale Analysis
Geometry
Tree Structure
Homogeneity
Data Structures
Interior
Pixel
Experiments
Verify
Output
Experiment

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Akbas, E., & Ahuja, N. (2010). From ramp discontinuities to segmentation tree. In Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers (PART 1 ed., pp. 123-134). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5994 LNCS, No. PART 1). https://doi.org/10.1007/978-3-642-12307-8_12

From ramp discontinuities to segmentation tree. / Akbas, Emre; Ahuja, Narendra.

Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers. PART 1. ed. 2010. p. 123-134 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5994 LNCS, No. PART 1).

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

Akbas, E & Ahuja, N 2010, From ramp discontinuities to segmentation tree. in Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 5994 LNCS, pp. 123-134, 9th Asian Conference on Computer Vision, ACCV 2009, Xi'an, China, 9/23/09. https://doi.org/10.1007/978-3-642-12307-8_12
Akbas E, Ahuja N. From ramp discontinuities to segmentation tree. In Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers. PART 1 ed. 2010. p. 123-134. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1). https://doi.org/10.1007/978-3-642-12307-8_12
Akbas, Emre ; Ahuja, Narendra. / From ramp discontinuities to segmentation tree. Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers. PART 1. ed. 2010. pp. 123-134 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).
@inproceedings{c771546f1ad44ad39c8114f2c88757c5,
title = "From ramp discontinuities to segmentation tree",
abstract = "This paper presents a new algorithm for low-level multiscale segmentation of images. The algorithm is designed to detect image regions regardless of their shapes, sizes, and levels of interior homogeneity, by doing a multiscale analysis without assuming any prior models of region geometry. As in previous work, a region is modeled as a homogeneous set of connected pixels surrounded by ramp discontinuities. A new transform, called the ramp transform, is described, which is used to detect ramp discontinuities and seeds for all regions in an image. Region seeds are grown towards the ramp discontinuity areas by utilizing a relaxation labeling procedure. Segmentation is achieved by analyzing the output of this procedure at multiple photometric scales. Finally, all detected regions are organized into a tree data structure based on their recursive containment relations. Experiments on real and synthetic images verify the desired properties of the proposed algorithm.",
author = "Emre Akbas and Narendra Ahuja",
year = "2010",
month = "12",
day = "29",
doi = "10.1007/978-3-642-12307-8_12",
language = "English (US)",
isbn = "3642123066",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 1",
pages = "123--134",
booktitle = "Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers",
edition = "PART 1",

}

TY - GEN

T1 - From ramp discontinuities to segmentation tree

AU - Akbas, Emre

AU - Ahuja, Narendra

PY - 2010/12/29

Y1 - 2010/12/29

N2 - This paper presents a new algorithm for low-level multiscale segmentation of images. The algorithm is designed to detect image regions regardless of their shapes, sizes, and levels of interior homogeneity, by doing a multiscale analysis without assuming any prior models of region geometry. As in previous work, a region is modeled as a homogeneous set of connected pixels surrounded by ramp discontinuities. A new transform, called the ramp transform, is described, which is used to detect ramp discontinuities and seeds for all regions in an image. Region seeds are grown towards the ramp discontinuity areas by utilizing a relaxation labeling procedure. Segmentation is achieved by analyzing the output of this procedure at multiple photometric scales. Finally, all detected regions are organized into a tree data structure based on their recursive containment relations. Experiments on real and synthetic images verify the desired properties of the proposed algorithm.

AB - This paper presents a new algorithm for low-level multiscale segmentation of images. The algorithm is designed to detect image regions regardless of their shapes, sizes, and levels of interior homogeneity, by doing a multiscale analysis without assuming any prior models of region geometry. As in previous work, a region is modeled as a homogeneous set of connected pixels surrounded by ramp discontinuities. A new transform, called the ramp transform, is described, which is used to detect ramp discontinuities and seeds for all regions in an image. Region seeds are grown towards the ramp discontinuity areas by utilizing a relaxation labeling procedure. Segmentation is achieved by analyzing the output of this procedure at multiple photometric scales. Finally, all detected regions are organized into a tree data structure based on their recursive containment relations. Experiments on real and synthetic images verify the desired properties of the proposed algorithm.

UR - http://www.scopus.com/inward/record.url?scp=78650465718&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78650465718&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-12307-8_12

DO - 10.1007/978-3-642-12307-8_12

M3 - Conference contribution

AN - SCOPUS:78650465718

SN - 3642123066

SN - 9783642123061

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

SP - 123

EP - 134

BT - Computer Vision, ACCV 2009 - 9th Asian Conference on Computer Vision, Revised Selected Papers

ER -