Finding objects by grouping primitives

David Forsyth, John Haddon, Sergey Ioffe

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

Abstract

Digital library applications require very general object recognition techniques. We describe an object recognition strategy that operates by grouping together image primitives in increasingly distinctive collections. Once a suciently large group has been found, we declare that an object is present. We demonstrate this method on applications such as nding unclothed people in general images and nding horses in general images. Finding clothed people is dicult, because the variation in colour and texture on the surface of clothing means that it is hard to nd regions of clothing in the image. We show that our strategy can be used to nd clothing by marking the distinctive shading patterns associated with folds in clothing, and then grouping these patterns.

Original languageEnglish (US)
Title of host publicationShape, Contour and Grouping in Computer Vision
EditorsJoseph L. Mundy, Roberto Cipolla, David A. Forsyth, Vito di Gesu
PublisherSpringer-Verlag
Pages302-318
Number of pages17
ISBN (Print)3540667229, 9783540667223
StatePublished - Jan 1 1999
EventInternational Workshop on Shape, Contour and Grouping in Computer Vision - Palermo, Sicily, Italy
Duration: May 26 1998May 29 1998

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1681
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

OtherInternational Workshop on Shape, Contour and Grouping in Computer Vision
CountryItaly
CityPalermo, Sicily
Period5/26/985/29/98

Fingerprint

Object recognition
Grouping
Digital libraries
Object Recognition
Textures
Color
Shading
Digital Libraries
Texture
Fold
Object
Demonstrate
Strategy

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Forsyth, D., Haddon, J., & Ioffe, S. (1999). Finding objects by grouping primitives. In J. L. Mundy, R. Cipolla, D. A. Forsyth, & V. di Gesu (Eds.), Shape, Contour and Grouping in Computer Vision (pp. 302-318). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1681). Springer-Verlag.

Finding objects by grouping primitives. / Forsyth, David; Haddon, John; Ioffe, Sergey.

Shape, Contour and Grouping in Computer Vision. ed. / Joseph L. Mundy; Roberto Cipolla; David A. Forsyth; Vito di Gesu. Springer-Verlag, 1999. p. 302-318 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1681).

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

Forsyth, D, Haddon, J & Ioffe, S 1999, Finding objects by grouping primitives. in JL Mundy, R Cipolla, DA Forsyth & V di Gesu (eds), Shape, Contour and Grouping in Computer Vision. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1681, Springer-Verlag, pp. 302-318, International Workshop on Shape, Contour and Grouping in Computer Vision, Palermo, Sicily, Italy, 5/26/98.
Forsyth D, Haddon J, Ioffe S. Finding objects by grouping primitives. In Mundy JL, Cipolla R, Forsyth DA, di Gesu V, editors, Shape, Contour and Grouping in Computer Vision. Springer-Verlag. 1999. p. 302-318. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Forsyth, David ; Haddon, John ; Ioffe, Sergey. / Finding objects by grouping primitives. Shape, Contour and Grouping in Computer Vision. editor / Joseph L. Mundy ; Roberto Cipolla ; David A. Forsyth ; Vito di Gesu. Springer-Verlag, 1999. pp. 302-318 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{16ed6b125a904f0791668c1047a251a3,
title = "Finding objects by grouping primitives",
abstract = "Digital library applications require very general object recognition techniques. We describe an object recognition strategy that operates by grouping together image primitives in increasingly distinctive collections. Once a suciently large group has been found, we declare that an object is present. We demonstrate this method on applications such as nding unclothed people in general images and nding horses in general images. Finding clothed people is dicult, because the variation in colour and texture on the surface of clothing means that it is hard to nd regions of clothing in the image. We show that our strategy can be used to nd clothing by marking the distinctive shading patterns associated with folds in clothing, and then grouping these patterns.",
author = "David Forsyth and John Haddon and Sergey Ioffe",
year = "1999",
month = "1",
day = "1",
language = "English (US)",
isbn = "3540667229",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer-Verlag",
pages = "302--318",
editor = "Mundy, {Joseph L.} and Roberto Cipolla and Forsyth, {David A.} and {di Gesu}, Vito",
booktitle = "Shape, Contour and Grouping in Computer Vision",

}

TY - GEN

T1 - Finding objects by grouping primitives

AU - Forsyth, David

AU - Haddon, John

AU - Ioffe, Sergey

PY - 1999/1/1

Y1 - 1999/1/1

N2 - Digital library applications require very general object recognition techniques. We describe an object recognition strategy that operates by grouping together image primitives in increasingly distinctive collections. Once a suciently large group has been found, we declare that an object is present. We demonstrate this method on applications such as nding unclothed people in general images and nding horses in general images. Finding clothed people is dicult, because the variation in colour and texture on the surface of clothing means that it is hard to nd regions of clothing in the image. We show that our strategy can be used to nd clothing by marking the distinctive shading patterns associated with folds in clothing, and then grouping these patterns.

AB - Digital library applications require very general object recognition techniques. We describe an object recognition strategy that operates by grouping together image primitives in increasingly distinctive collections. Once a suciently large group has been found, we declare that an object is present. We demonstrate this method on applications such as nding unclothed people in general images and nding horses in general images. Finding clothed people is dicult, because the variation in colour and texture on the surface of clothing means that it is hard to nd regions of clothing in the image. We show that our strategy can be used to nd clothing by marking the distinctive shading patterns associated with folds in clothing, and then grouping these patterns.

UR - http://www.scopus.com/inward/record.url?scp=34249055330&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34249055330&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:34249055330

SN - 3540667229

SN - 9783540667223

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

SP - 302

EP - 318

BT - Shape, Contour and Grouping in Computer Vision

A2 - Mundy, Joseph L.

A2 - Cipolla, Roberto

A2 - Forsyth, David A.

A2 - di Gesu, Vito

PB - Springer-Verlag

ER -