Eigenwalks

Walk detection and biometrics from symmetry patterns

László Havasi, Zoltán Szlávik, T. Szirányi

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

11 Citations (Scopus)

Abstract

In this paper we present a symmetry-based approach which can be used to detect humans and to extract biometric characteristics from video image-sequences. The method employs a simplified symmetry-feature extracted from the images. To obtain a useful descriptor of a walking person, we track temporally the symmetries which result from the movements of the person's legs. In a further processing stage these patterns are filtered, then re-sampled using Bezier-splines to generate an invariant and noise-cleaned signature or "feature". In our detection method the extracted spatio-temporal feature with a large number of dimensions (800) is transformed to a space with a much smaller number of dimensions (3), which we call the "eigenwalks space"; the method uses Principal Component Analysis (PCA) to reduce the dimensionality, and the Support Vector Machine (SVM) method in the eigenspace for recognition purposes. Finally we present a method by which we can estimate the gait-parameters (the beginning and end of a walk-cycle, identification of the leading leg) from the symmetry-patterns of the walking person, without camera calibration, based on two successive detected walk-steps.

Original languageEnglish
Title of host publicationProceedings - International Conference on Image Processing, ICIP
Pages289-292
Number of pages4
Volume3
DOIs
Publication statusPublished - 2005
EventIEEE International Conference on Image Processing 2005, ICIP 2005 - Genova, Italy
Duration: Sep 11 2005Sep 14 2005

Other

OtherIEEE International Conference on Image Processing 2005, ICIP 2005
CountryItaly
CityGenova
Period9/11/059/14/05

Fingerprint

Biometrics
Splines
Principal component analysis
Support vector machines
Cameras
Calibration
Processing

Keywords

  • Component
  • Motion analysis
  • PCA
  • Pedestrian detection
  • Spline interpolation
  • SVM

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Havasi, L., Szlávik, Z., & Szirányi, T. (2005). Eigenwalks: Walk detection and biometrics from symmetry patterns. In Proceedings - International Conference on Image Processing, ICIP (Vol. 3, pp. 289-292). [1530385] https://doi.org/10.1109/ICIP.2005.1530385

Eigenwalks : Walk detection and biometrics from symmetry patterns. / Havasi, László; Szlávik, Zoltán; Szirányi, T.

Proceedings - International Conference on Image Processing, ICIP. Vol. 3 2005. p. 289-292 1530385.

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

Havasi, L, Szlávik, Z & Szirányi, T 2005, Eigenwalks: Walk detection and biometrics from symmetry patterns. in Proceedings - International Conference on Image Processing, ICIP. vol. 3, 1530385, pp. 289-292, IEEE International Conference on Image Processing 2005, ICIP 2005, Genova, Italy, 9/11/05. https://doi.org/10.1109/ICIP.2005.1530385
Havasi L, Szlávik Z, Szirányi T. Eigenwalks: Walk detection and biometrics from symmetry patterns. In Proceedings - International Conference on Image Processing, ICIP. Vol. 3. 2005. p. 289-292. 1530385 https://doi.org/10.1109/ICIP.2005.1530385
Havasi, László ; Szlávik, Zoltán ; Szirányi, T. / Eigenwalks : Walk detection and biometrics from symmetry patterns. Proceedings - International Conference on Image Processing, ICIP. Vol. 3 2005. pp. 289-292
@inproceedings{b73f18133ad14971b83bb7147974d321,
title = "Eigenwalks: Walk detection and biometrics from symmetry patterns",
abstract = "In this paper we present a symmetry-based approach which can be used to detect humans and to extract biometric characteristics from video image-sequences. The method employs a simplified symmetry-feature extracted from the images. To obtain a useful descriptor of a walking person, we track temporally the symmetries which result from the movements of the person's legs. In a further processing stage these patterns are filtered, then re-sampled using Bezier-splines to generate an invariant and noise-cleaned signature or {"}feature{"}. In our detection method the extracted spatio-temporal feature with a large number of dimensions (800) is transformed to a space with a much smaller number of dimensions (3), which we call the {"}eigenwalks space{"}; the method uses Principal Component Analysis (PCA) to reduce the dimensionality, and the Support Vector Machine (SVM) method in the eigenspace for recognition purposes. Finally we present a method by which we can estimate the gait-parameters (the beginning and end of a walk-cycle, identification of the leading leg) from the symmetry-patterns of the walking person, without camera calibration, based on two successive detected walk-steps.",
keywords = "Component, Motion analysis, PCA, Pedestrian detection, Spline interpolation, SVM",
author = "L{\'a}szl{\'o} Havasi and Zolt{\'a}n Szl{\'a}vik and T. Szir{\'a}nyi",
year = "2005",
doi = "10.1109/ICIP.2005.1530385",
language = "English",
isbn = "0780391349",
volume = "3",
pages = "289--292",
booktitle = "Proceedings - International Conference on Image Processing, ICIP",

}

TY - GEN

T1 - Eigenwalks

T2 - Walk detection and biometrics from symmetry patterns

AU - Havasi, László

AU - Szlávik, Zoltán

AU - Szirányi, T.

PY - 2005

Y1 - 2005

N2 - In this paper we present a symmetry-based approach which can be used to detect humans and to extract biometric characteristics from video image-sequences. The method employs a simplified symmetry-feature extracted from the images. To obtain a useful descriptor of a walking person, we track temporally the symmetries which result from the movements of the person's legs. In a further processing stage these patterns are filtered, then re-sampled using Bezier-splines to generate an invariant and noise-cleaned signature or "feature". In our detection method the extracted spatio-temporal feature with a large number of dimensions (800) is transformed to a space with a much smaller number of dimensions (3), which we call the "eigenwalks space"; the method uses Principal Component Analysis (PCA) to reduce the dimensionality, and the Support Vector Machine (SVM) method in the eigenspace for recognition purposes. Finally we present a method by which we can estimate the gait-parameters (the beginning and end of a walk-cycle, identification of the leading leg) from the symmetry-patterns of the walking person, without camera calibration, based on two successive detected walk-steps.

AB - In this paper we present a symmetry-based approach which can be used to detect humans and to extract biometric characteristics from video image-sequences. The method employs a simplified symmetry-feature extracted from the images. To obtain a useful descriptor of a walking person, we track temporally the symmetries which result from the movements of the person's legs. In a further processing stage these patterns are filtered, then re-sampled using Bezier-splines to generate an invariant and noise-cleaned signature or "feature". In our detection method the extracted spatio-temporal feature with a large number of dimensions (800) is transformed to a space with a much smaller number of dimensions (3), which we call the "eigenwalks space"; the method uses Principal Component Analysis (PCA) to reduce the dimensionality, and the Support Vector Machine (SVM) method in the eigenspace for recognition purposes. Finally we present a method by which we can estimate the gait-parameters (the beginning and end of a walk-cycle, identification of the leading leg) from the symmetry-patterns of the walking person, without camera calibration, based on two successive detected walk-steps.

KW - Component

KW - Motion analysis

KW - PCA

KW - Pedestrian detection

KW - Spline interpolation

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=33749246056&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33749246056&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2005.1530385

DO - 10.1109/ICIP.2005.1530385

M3 - Conference contribution

SN - 0780391349

SN - 9780780391345

VL - 3

SP - 289

EP - 292

BT - Proceedings - International Conference on Image Processing, ICIP

ER -