An algorithm for recognising walkers

H. M. Lakany, G. M. Hayes

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

5 Citations (Scopus)

Abstract

In this paper, we present an algorithm to recognise walking people, based upon extracting the spatio-temporal trajectories of the joints of a walking subject.

Subjects are filmed with LEDs attached to their joints and head such that the lights are the only objects visible in the film sequence — a method known as moving light displays (MLDs). Lights are tracked through the sequence of frames and are labelled based on human walking behaviour. In the case of self-occluded lights, a radial basis function neural network was trained and used for predicting the positions of occluded markers. The trajectory of each MLD is transformed using a 2D fast Fourier transform. Components of the FFT for all MLDs are considered as the feature vector of each subject. This is fed to a multi-layer perceptron (MLP) for classification.

The algorithm was used to recognise four subjects — 3 males and 1 female. For each subject, 10 gait cycles were used for training and 5 for testing the MLP. Backpropagation was used to train the network. Results show that the algorithm is a promising technique for recognising subjects by their gait.
LanguageEnglish
Title of host publicationInternational Conference on Audio- and Video-Based Biometric Person Authentication
Subtitle of host publicationFirst International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings
PublisherSpringer
Pages111-118
Number of pages8
ISBN (Electronic)978-3-540-68425-1
ISBN (Print)978-3-540-62660-2
DOIs
Publication statusPublished - 1997

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume1206
ISSN (Print)0302-9743

Fingerprint

Display devices
Multilayer neural networks
Fast Fourier transforms
Trajectories
Backpropagation
Light emitting diodes
Neural networks
Testing

Keywords

  • spatio-temporal trajectories
  • walking
  • gait cycles
  • gait recognition

Cite this

Lakany, H. M., & Hayes, G. M. (1997). An algorithm for recognising walkers. In International Conference on Audio- and Video-Based Biometric Person Authentication: First International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings (pp. 111-118). (Lecture Notes in Computer Science ; Vol. 1206). Springer. https://doi.org/10.1007/BFb0015986
Lakany, H. M. ; Hayes, G. M. . / An algorithm for recognising walkers. International Conference on Audio- and Video-Based Biometric Person Authentication: First International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings. Springer, 1997. pp. 111-118 (Lecture Notes in Computer Science ).
@inproceedings{856368ff49ac4a479d7c30cdc285233a,
title = "An algorithm for recognising walkers",
abstract = "In this paper, we present an algorithm to recognise walking people, based upon extracting the spatio-temporal trajectories of the joints of a walking subject.Subjects are filmed with LEDs attached to their joints and head such that the lights are the only objects visible in the film sequence — a method known as moving light displays (MLDs). Lights are tracked through the sequence of frames and are labelled based on human walking behaviour. In the case of self-occluded lights, a radial basis function neural network was trained and used for predicting the positions of occluded markers. The trajectory of each MLD is transformed using a 2D fast Fourier transform. Components of the FFT for all MLDs are considered as the feature vector of each subject. This is fed to a multi-layer perceptron (MLP) for classification.The algorithm was used to recognise four subjects — 3 males and 1 female. For each subject, 10 gait cycles were used for training and 5 for testing the MLP. Backpropagation was used to train the network. Results show that the algorithm is a promising technique for recognising subjects by their gait.",
keywords = "spatio-temporal trajectories, walking, gait cycles, gait recognition",
author = "Lakany, {H. M.} and Hayes, {G. M.}",
year = "1997",
doi = "10.1007/BFb0015986",
language = "English",
isbn = "978-3-540-62660-2",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "111--118",
booktitle = "International Conference on Audio- and Video-Based Biometric Person Authentication",

}

Lakany, HM & Hayes, GM 1997, An algorithm for recognising walkers. in International Conference on Audio- and Video-Based Biometric Person Authentication: First International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings. Lecture Notes in Computer Science , vol. 1206, Springer, pp. 111-118. https://doi.org/10.1007/BFb0015986

An algorithm for recognising walkers. / Lakany, H. M.; Hayes, G. M. .

International Conference on Audio- and Video-Based Biometric Person Authentication: First International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings. Springer, 1997. p. 111-118 (Lecture Notes in Computer Science ; Vol. 1206).

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

TY - GEN

T1 - An algorithm for recognising walkers

AU - Lakany, H. M.

AU - Hayes, G. M.

PY - 1997

Y1 - 1997

N2 - In this paper, we present an algorithm to recognise walking people, based upon extracting the spatio-temporal trajectories of the joints of a walking subject.Subjects are filmed with LEDs attached to their joints and head such that the lights are the only objects visible in the film sequence — a method known as moving light displays (MLDs). Lights are tracked through the sequence of frames and are labelled based on human walking behaviour. In the case of self-occluded lights, a radial basis function neural network was trained and used for predicting the positions of occluded markers. The trajectory of each MLD is transformed using a 2D fast Fourier transform. Components of the FFT for all MLDs are considered as the feature vector of each subject. This is fed to a multi-layer perceptron (MLP) for classification.The algorithm was used to recognise four subjects — 3 males and 1 female. For each subject, 10 gait cycles were used for training and 5 for testing the MLP. Backpropagation was used to train the network. Results show that the algorithm is a promising technique for recognising subjects by their gait.

AB - In this paper, we present an algorithm to recognise walking people, based upon extracting the spatio-temporal trajectories of the joints of a walking subject.Subjects are filmed with LEDs attached to their joints and head such that the lights are the only objects visible in the film sequence — a method known as moving light displays (MLDs). Lights are tracked through the sequence of frames and are labelled based on human walking behaviour. In the case of self-occluded lights, a radial basis function neural network was trained and used for predicting the positions of occluded markers. The trajectory of each MLD is transformed using a 2D fast Fourier transform. Components of the FFT for all MLDs are considered as the feature vector of each subject. This is fed to a multi-layer perceptron (MLP) for classification.The algorithm was used to recognise four subjects — 3 males and 1 female. For each subject, 10 gait cycles were used for training and 5 for testing the MLP. Backpropagation was used to train the network. Results show that the algorithm is a promising technique for recognising subjects by their gait.

KW - spatio-temporal trajectories

KW - walking

KW - gait cycles

KW - gait recognition

U2 - 10.1007/BFb0015986

DO - 10.1007/BFb0015986

M3 - Conference contribution book

SN - 978-3-540-62660-2

T3 - Lecture Notes in Computer Science

SP - 111

EP - 118

BT - International Conference on Audio- and Video-Based Biometric Person Authentication

PB - Springer

ER -

Lakany HM, Hayes GM. An algorithm for recognising walkers. In International Conference on Audio- and Video-Based Biometric Person Authentication: First International Conference, AVBPA '97, Crans-Montana, Switzerland, March 12 - 14, 1997, Proceedings. Springer. 1997. p. 111-118. (Lecture Notes in Computer Science ). https://doi.org/10.1007/BFb0015986