Junctions

detection, classification, and reconstruction

Laxmi Parida, Davi Geiger, Robert Hummel

Research output: Contribution to journalArticle

Abstract

Junctions are important features for image analysis and form a critical aspect of image understanding tasks such as object recognition. We present a unified approach to detecting (location of the center of the junction), classifying (by the number of wedges-lines, corners, three-junctions such as T or Y junctions, or four-junctions such as X-junctions), and reconstructing junctions (in terms of radius size, the angles of each wedge and the intensity in each of the wedges) in images. Our main contribution is a modeling of the junction which is complex enough to handle all these issues and yet simple enough to admit an effective dynamic programming solution. Broadly, we use a template deformation framework along with a gradient criterium to detect radial partitions of the template. We use the minimum description length principle to obtain the optimal number of partitions that best describes the junction. Kona [27] is an implementation of this model. We (quantitatively) demonstrate the stability and robustness of the detector by analyzing its behavior in the presence of noise, using synthetic/controlled apparatus. We also present a qualitative study of its behavior on real images.

Original languageEnglish (US)
Pages (from-to)687-698
Number of pages12
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Volume20
Issue number7
DOIs
StatePublished - 1998

Fingerprint

Image understanding
Object recognition
Wedge
Dynamic programming
Image analysis
Detectors
Template
Partition
Image Understanding
Object Recognition
Image Analysis
Dynamic Programming
Radius
Detector
Gradient
Robustness
Angle
Line
Modeling
Demonstrate

Keywords

  • Corners
  • Energy minimization. © 1998 ieee
  • Feature detection
  • Junctions
  • Low-level vision
  • Minimum description length (MDL) principle

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition

Cite this

Junctions : detection, classification, and reconstruction. / Parida, Laxmi; Geiger, Davi; Hummel, Robert.

In: IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 20, No. 7, 1998, p. 687-698.

Research output: Contribution to journalArticle

@article{9960f1ccfa294d7dbf5be7c6efbf92ae,
title = "Junctions: detection, classification, and reconstruction",
abstract = "Junctions are important features for image analysis and form a critical aspect of image understanding tasks such as object recognition. We present a unified approach to detecting (location of the center of the junction), classifying (by the number of wedges-lines, corners, three-junctions such as T or Y junctions, or four-junctions such as X-junctions), and reconstructing junctions (in terms of radius size, the angles of each wedge and the intensity in each of the wedges) in images. Our main contribution is a modeling of the junction which is complex enough to handle all these issues and yet simple enough to admit an effective dynamic programming solution. Broadly, we use a template deformation framework along with a gradient criterium to detect radial partitions of the template. We use the minimum description length principle to obtain the optimal number of partitions that best describes the junction. Kona [27] is an implementation of this model. We (quantitatively) demonstrate the stability and robustness of the detector by analyzing its behavior in the presence of noise, using synthetic/controlled apparatus. We also present a qualitative study of its behavior on real images.",
keywords = "Corners, Energy minimization. {\circledC} 1998 ieee, Feature detection, Junctions, Low-level vision, Minimum description length (MDL) principle",
author = "Laxmi Parida and Davi Geiger and Robert Hummel",
year = "1998",
doi = "10.1109/34.689300",
language = "English (US)",
volume = "20",
pages = "687--698",
journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",
issn = "0162-8828",
publisher = "IEEE Computer Society",
number = "7",

}

TY - JOUR

T1 - Junctions

T2 - detection, classification, and reconstruction

AU - Parida, Laxmi

AU - Geiger, Davi

AU - Hummel, Robert

PY - 1998

Y1 - 1998

N2 - Junctions are important features for image analysis and form a critical aspect of image understanding tasks such as object recognition. We present a unified approach to detecting (location of the center of the junction), classifying (by the number of wedges-lines, corners, three-junctions such as T or Y junctions, or four-junctions such as X-junctions), and reconstructing junctions (in terms of radius size, the angles of each wedge and the intensity in each of the wedges) in images. Our main contribution is a modeling of the junction which is complex enough to handle all these issues and yet simple enough to admit an effective dynamic programming solution. Broadly, we use a template deformation framework along with a gradient criterium to detect radial partitions of the template. We use the minimum description length principle to obtain the optimal number of partitions that best describes the junction. Kona [27] is an implementation of this model. We (quantitatively) demonstrate the stability and robustness of the detector by analyzing its behavior in the presence of noise, using synthetic/controlled apparatus. We also present a qualitative study of its behavior on real images.

AB - Junctions are important features for image analysis and form a critical aspect of image understanding tasks such as object recognition. We present a unified approach to detecting (location of the center of the junction), classifying (by the number of wedges-lines, corners, three-junctions such as T or Y junctions, or four-junctions such as X-junctions), and reconstructing junctions (in terms of radius size, the angles of each wedge and the intensity in each of the wedges) in images. Our main contribution is a modeling of the junction which is complex enough to handle all these issues and yet simple enough to admit an effective dynamic programming solution. Broadly, we use a template deformation framework along with a gradient criterium to detect radial partitions of the template. We use the minimum description length principle to obtain the optimal number of partitions that best describes the junction. Kona [27] is an implementation of this model. We (quantitatively) demonstrate the stability and robustness of the detector by analyzing its behavior in the presence of noise, using synthetic/controlled apparatus. We also present a qualitative study of its behavior on real images.

KW - Corners

KW - Energy minimization. © 1998 ieee

KW - Feature detection

KW - Junctions

KW - Low-level vision

KW - Minimum description length (MDL) principle

UR - http://www.scopus.com/inward/record.url?scp=0032124295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032124295&partnerID=8YFLogxK

U2 - 10.1109/34.689300

DO - 10.1109/34.689300

M3 - Article

VL - 20

SP - 687

EP - 698

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

SN - 0162-8828

IS - 7

ER -