Kona: A multi-junction detector using minimum description length principle

Laxmi Parida, Davi Geiger, Robert Hummel

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

Abstract

Corners, T-, Y-, X-junctions give vital depth cues which is a critical aspect of image understanding tasks like object recognition: junctions form an important class of features invaluable in most vision systems. The three main issues in a junction (or any feature) detector are: scale, location, and, the junction (feature) parameters. The junction parameters are (1) the radius, or size, of the junction, (2) the kind of junction: lines, corners, 3-junctions such as T or Y, or, 4-junction such as X-junction, etcetera, (3) angles of the wedges, and, (4) intensity in each of the wedges. Our main contribution in this paper is a modeling of the junction (using the minimum description length principle), which is complex enough to handle all the three issues and simple enough to admit an effective dynamic programming solution. Kona is an implementation of this model. A similar approach can be used to model other features like thick edges, blobs and end-points.

Original languageEnglish (US)
Title of host publicationEnergy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings
PublisherSpringer Verlag
Pages51-65
Number of pages15
Volume1223
ISBN (Print)3540629092, 9783540629092
DOIs
StatePublished - 1997
EventInternational Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997 - Venice, Italy
Duration: May 21 1997May 23 1997

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1223
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

OtherInternational Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997
CountryItaly
CityVenice
Period5/21/975/23/97

Fingerprint

Wedge
Detector
Detectors
Image Understanding
Image understanding
Object recognition
Object Recognition
Vision System
End point
Dynamic programming
Dynamic Programming
Radius
Angle
Line
Modeling
Model
Class

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Parida, L., Geiger, D., & Hummel, R. (1997). Kona: A multi-junction detector using minimum description length principle. In Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings (Vol. 1223, pp. 51-65). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1223). Springer Verlag. https://doi.org/10.1007/3-540-62909-2_72

Kona : A multi-junction detector using minimum description length principle. / Parida, Laxmi; Geiger, Davi; Hummel, Robert.

Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223 Springer Verlag, 1997. p. 51-65 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1223).

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

Parida, L, Geiger, D & Hummel, R 1997, Kona: A multi-junction detector using minimum description length principle. in Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. vol. 1223, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1223, Springer Verlag, pp. 51-65, International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 1997, Venice, Italy, 5/21/97. https://doi.org/10.1007/3-540-62909-2_72
Parida L, Geiger D, Hummel R. Kona: A multi-junction detector using minimum description length principle. In Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223. Springer Verlag. 1997. p. 51-65. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-62909-2_72
Parida, Laxmi ; Geiger, Davi ; Hummel, Robert. / Kona : A multi-junction detector using minimum description length principle. Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings. Vol. 1223 Springer Verlag, 1997. pp. 51-65 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ca0edcce2e4b4579813e1640e5cfa70d,
title = "Kona: A multi-junction detector using minimum description length principle",
abstract = "Corners, T-, Y-, X-junctions give vital depth cues which is a critical aspect of image understanding tasks like object recognition: junctions form an important class of features invaluable in most vision systems. The three main issues in a junction (or any feature) detector are: scale, location, and, the junction (feature) parameters. The junction parameters are (1) the radius, or size, of the junction, (2) the kind of junction: lines, corners, 3-junctions such as T or Y, or, 4-junction such as X-junction, etcetera, (3) angles of the wedges, and, (4) intensity in each of the wedges. Our main contribution in this paper is a modeling of the junction (using the minimum description length principle), which is complex enough to handle all the three issues and simple enough to admit an effective dynamic programming solution. Kona is an implementation of this model. A similar approach can be used to model other features like thick edges, blobs and end-points.",
author = "Laxmi Parida and Davi Geiger and Robert Hummel",
year = "1997",
doi = "10.1007/3-540-62909-2_72",
language = "English (US)",
isbn = "3540629092",
volume = "1223",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "51--65",
booktitle = "Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings",

}

TY - GEN

T1 - Kona

T2 - A multi-junction detector using minimum description length principle

AU - Parida, Laxmi

AU - Geiger, Davi

AU - Hummel, Robert

PY - 1997

Y1 - 1997

N2 - Corners, T-, Y-, X-junctions give vital depth cues which is a critical aspect of image understanding tasks like object recognition: junctions form an important class of features invaluable in most vision systems. The three main issues in a junction (or any feature) detector are: scale, location, and, the junction (feature) parameters. The junction parameters are (1) the radius, or size, of the junction, (2) the kind of junction: lines, corners, 3-junctions such as T or Y, or, 4-junction such as X-junction, etcetera, (3) angles of the wedges, and, (4) intensity in each of the wedges. Our main contribution in this paper is a modeling of the junction (using the minimum description length principle), which is complex enough to handle all the three issues and simple enough to admit an effective dynamic programming solution. Kona is an implementation of this model. A similar approach can be used to model other features like thick edges, blobs and end-points.

AB - Corners, T-, Y-, X-junctions give vital depth cues which is a critical aspect of image understanding tasks like object recognition: junctions form an important class of features invaluable in most vision systems. The three main issues in a junction (or any feature) detector are: scale, location, and, the junction (feature) parameters. The junction parameters are (1) the radius, or size, of the junction, (2) the kind of junction: lines, corners, 3-junctions such as T or Y, or, 4-junction such as X-junction, etcetera, (3) angles of the wedges, and, (4) intensity in each of the wedges. Our main contribution in this paper is a modeling of the junction (using the minimum description length principle), which is complex enough to handle all the three issues and simple enough to admit an effective dynamic programming solution. Kona is an implementation of this model. A similar approach can be used to model other features like thick edges, blobs and end-points.

UR - http://www.scopus.com/inward/record.url?scp=33747805205&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33747805205&partnerID=8YFLogxK

U2 - 10.1007/3-540-62909-2_72

DO - 10.1007/3-540-62909-2_72

M3 - Conference contribution

AN - SCOPUS:33747805205

SN - 3540629092

SN - 9783540629092

VL - 1223

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

SP - 51

EP - 65

BT - Energy Minimization Methods in Computer Vision and Pattern Recognition - International Workshop EMMCVPR 1997, Proceedings

PB - Springer Verlag

ER -