A continuous shape descriptor by orientation diffusion

Hsing Kuo Pao, Davi Geiger

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

Abstract

We propose a continuous description for 2-D shapes that calculates convexity, symmetry and is able to account for size. Convexity and size are known to be critical in deciding figure/ground (F/G) separation, with the study initiated by the Gestalt school [9] [11]. However, few quantitative discussions were made before. Thus, we emphasize the convexity/size measurement for the purpose of F/G prediction. A Kullback-Leibler measure is introduced. In addition, the symmetry information is studied through the same platform. All these shape properties are collected for shape representations. Overall, our representations are given in a continuous manner. For convexity measurement, unlike the 1/0 mathematical definition where shapes are categorized as convex or concave, we give a measure describing shapes as “more” or “less” convex than others. In symmetry information (skeleton) retrieval, a 2-D intensity map is provided with the intensity value specifying “strength” of the skeleton. The proposed representations are robust in the sense that small fine-scale perturbations on shape boundaries will cause minor effects on the final representations. All these shape properties are intergrated into one description. To apply to the F/G separation, the shape measure can be flexibly chosen between a size-invariant convexity measure or a convexity measure with the small size preference. The model is established on an orientation diffusion framework, where the local features, served as inputs, are intensity edge locations and their orientations. The approach is a variational one, rooted in a Markov random field (MRF) formulation. A quadratic form is used to assure simplicity and the existence of solution.

Original languageEnglish (US)
Title of host publicationEnergy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings
PublisherSpringer Verlag
Pages544-559
Number of pages16
Volume2134
ISBN (Print)3540425233, 9783540425236
DOIs
StatePublished - 2001
Event3rd International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 2001 - Sophia Antipolis, France
Duration: Sep 3 2001Sep 5 2001

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2134
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other3rd International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 2001
CountryFrance
CitySophia Antipolis
Period9/3/019/5/01

Fingerprint

Shape Descriptor
Convexity
Information retrieval
Figure
Skeleton
Symmetry
Shape Representation
Local Features
Quadratic form
Random Field
Existence of Solutions
Minor
Simplicity
Retrieval
Perturbation
Calculate
Invariant
Formulation
Prediction

Keywords

  • Convexity
  • Early vision
  • Orientation diffusion
  • Shape analysis
  • Symmetry

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Pao, H. K., & Geiger, D. (2001). A continuous shape descriptor by orientation diffusion. In Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings (Vol. 2134, pp. 544-559). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2134). Springer Verlag. https://doi.org/10.1007/3-540-44745-8_36

A continuous shape descriptor by orientation diffusion. / Pao, Hsing Kuo; Geiger, Davi.

Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings. Vol. 2134 Springer Verlag, 2001. p. 544-559 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2134).

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

Pao, HK & Geiger, D 2001, A continuous shape descriptor by orientation diffusion. in Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings. vol. 2134, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2134, Springer Verlag, pp. 544-559, 3rd International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition, EMMCVPR 2001, Sophia Antipolis, France, 9/3/01. https://doi.org/10.1007/3-540-44745-8_36
Pao HK, Geiger D. A continuous shape descriptor by orientation diffusion. In Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings. Vol. 2134. Springer Verlag. 2001. p. 544-559. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-44745-8_36
Pao, Hsing Kuo ; Geiger, Davi. / A continuous shape descriptor by orientation diffusion. Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings. Vol. 2134 Springer Verlag, 2001. pp. 544-559 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{7ce9f979719f4f3a82c24e1b4015a90f,
title = "A continuous shape descriptor by orientation diffusion",
abstract = "We propose a continuous description for 2-D shapes that calculates convexity, symmetry and is able to account for size. Convexity and size are known to be critical in deciding figure/ground (F/G) separation, with the study initiated by the Gestalt school [9] [11]. However, few quantitative discussions were made before. Thus, we emphasize the convexity/size measurement for the purpose of F/G prediction. A Kullback-Leibler measure is introduced. In addition, the symmetry information is studied through the same platform. All these shape properties are collected for shape representations. Overall, our representations are given in a continuous manner. For convexity measurement, unlike the 1/0 mathematical definition where shapes are categorized as convex or concave, we give a measure describing shapes as “more” or “less” convex than others. In symmetry information (skeleton) retrieval, a 2-D intensity map is provided with the intensity value specifying “strength” of the skeleton. The proposed representations are robust in the sense that small fine-scale perturbations on shape boundaries will cause minor effects on the final representations. All these shape properties are intergrated into one description. To apply to the F/G separation, the shape measure can be flexibly chosen between a size-invariant convexity measure or a convexity measure with the small size preference. The model is established on an orientation diffusion framework, where the local features, served as inputs, are intensity edge locations and their orientations. The approach is a variational one, rooted in a Markov random field (MRF) formulation. A quadratic form is used to assure simplicity and the existence of solution.",
keywords = "Convexity, Early vision, Orientation diffusion, Shape analysis, Symmetry",
author = "Pao, {Hsing Kuo} and Davi Geiger",
year = "2001",
doi = "10.1007/3-540-44745-8_36",
language = "English (US)",
isbn = "3540425233",
volume = "2134",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "544--559",
booktitle = "Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings",

}

TY - GEN

T1 - A continuous shape descriptor by orientation diffusion

AU - Pao, Hsing Kuo

AU - Geiger, Davi

PY - 2001

Y1 - 2001

N2 - We propose a continuous description for 2-D shapes that calculates convexity, symmetry and is able to account for size. Convexity and size are known to be critical in deciding figure/ground (F/G) separation, with the study initiated by the Gestalt school [9] [11]. However, few quantitative discussions were made before. Thus, we emphasize the convexity/size measurement for the purpose of F/G prediction. A Kullback-Leibler measure is introduced. In addition, the symmetry information is studied through the same platform. All these shape properties are collected for shape representations. Overall, our representations are given in a continuous manner. For convexity measurement, unlike the 1/0 mathematical definition where shapes are categorized as convex or concave, we give a measure describing shapes as “more” or “less” convex than others. In symmetry information (skeleton) retrieval, a 2-D intensity map is provided with the intensity value specifying “strength” of the skeleton. The proposed representations are robust in the sense that small fine-scale perturbations on shape boundaries will cause minor effects on the final representations. All these shape properties are intergrated into one description. To apply to the F/G separation, the shape measure can be flexibly chosen between a size-invariant convexity measure or a convexity measure with the small size preference. The model is established on an orientation diffusion framework, where the local features, served as inputs, are intensity edge locations and their orientations. The approach is a variational one, rooted in a Markov random field (MRF) formulation. A quadratic form is used to assure simplicity and the existence of solution.

AB - We propose a continuous description for 2-D shapes that calculates convexity, symmetry and is able to account for size. Convexity and size are known to be critical in deciding figure/ground (F/G) separation, with the study initiated by the Gestalt school [9] [11]. However, few quantitative discussions were made before. Thus, we emphasize the convexity/size measurement for the purpose of F/G prediction. A Kullback-Leibler measure is introduced. In addition, the symmetry information is studied through the same platform. All these shape properties are collected for shape representations. Overall, our representations are given in a continuous manner. For convexity measurement, unlike the 1/0 mathematical definition where shapes are categorized as convex or concave, we give a measure describing shapes as “more” or “less” convex than others. In symmetry information (skeleton) retrieval, a 2-D intensity map is provided with the intensity value specifying “strength” of the skeleton. The proposed representations are robust in the sense that small fine-scale perturbations on shape boundaries will cause minor effects on the final representations. All these shape properties are intergrated into one description. To apply to the F/G separation, the shape measure can be flexibly chosen between a size-invariant convexity measure or a convexity measure with the small size preference. The model is established on an orientation diffusion framework, where the local features, served as inputs, are intensity edge locations and their orientations. The approach is a variational one, rooted in a Markov random field (MRF) formulation. A quadratic form is used to assure simplicity and the existence of solution.

KW - Convexity

KW - Early vision

KW - Orientation diffusion

KW - Shape analysis

KW - Symmetry

UR - http://www.scopus.com/inward/record.url?scp=33748451444&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748451444&partnerID=8YFLogxK

U2 - 10.1007/3-540-44745-8_36

DO - 10.1007/3-540-44745-8_36

M3 - Conference contribution

SN - 3540425233

SN - 9783540425236

VL - 2134

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

SP - 544

EP - 559

BT - Energy Minimization Methods in Computer Vision and Pattern Recognition - 3rd International Workshop, EMMCVPR 2001, Proceedings

PB - Springer Verlag

ER -