Vision-based recognition of hand shapes in taiwanese sign language

Jung Ning Huang, Pi Fuei Hsieh, Chung-Hsien Wu

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

Abstract

The pixel-based shape representation has been sensitive to rotation. In this paper, we propose a pixel-based descriptor that is invariant with rotation and scale for the hand shape recognition in Taiwanese Sign Language (TSL). Based on the property that a hand shape is characteristic of a unique pointing direction, angle normalization is used to meet the rotation-invariant requirement. With angle normalization, the traces of class covariance matrices have been reduced almost all over the classes of hand shapes, implying a less overlap between classes. It is confirmed by the experiments that show an increase in recognition accuracy.

Original languageEnglish
Title of host publicationAffective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings
Pages224-231
Number of pages8
Volume3784 LNCS
Publication statusPublished - 2005
Event1st International Conference on ffective Computing and Intelligent Interaction, ACII 2005 - Beijing, China
Duration: 2005 Oct 222005 Oct 24

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3784 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other1st International Conference on ffective Computing and Intelligent Interaction, ACII 2005
CountryChina
CityBeijing
Period05-10-2205-10-24

Fingerprint

Sign Language
Hand
Normalization
Pixel
Pixels
Direction angles
Shape Recognition
Shape Representation
Rotation Invariant
Covariance matrix
Descriptors
Overlap
Trace
Angle
Invariant
Requirements
Experiment
Vision
Class
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

Huang, J. N., Hsieh, P. F., & Wu, C-H. (2005). Vision-based recognition of hand shapes in taiwanese sign language. In Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings (Vol. 3784 LNCS, pp. 224-231). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3784 LNCS).
Huang, Jung Ning ; Hsieh, Pi Fuei ; Wu, Chung-Hsien. / Vision-based recognition of hand shapes in taiwanese sign language. Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings. Vol. 3784 LNCS 2005. pp. 224-231 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{dcf49a5df3f244cdbb1458694e354dd1,
title = "Vision-based recognition of hand shapes in taiwanese sign language",
abstract = "The pixel-based shape representation has been sensitive to rotation. In this paper, we propose a pixel-based descriptor that is invariant with rotation and scale for the hand shape recognition in Taiwanese Sign Language (TSL). Based on the property that a hand shape is characteristic of a unique pointing direction, angle normalization is used to meet the rotation-invariant requirement. With angle normalization, the traces of class covariance matrices have been reduced almost all over the classes of hand shapes, implying a less overlap between classes. It is confirmed by the experiments that show an increase in recognition accuracy.",
author = "Huang, {Jung Ning} and Hsieh, {Pi Fuei} and Chung-Hsien Wu",
year = "2005",
language = "English",
isbn = "3540296212",
volume = "3784 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "224--231",
booktitle = "Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings",

}

Huang, JN, Hsieh, PF & Wu, C-H 2005, Vision-based recognition of hand shapes in taiwanese sign language. in Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings. vol. 3784 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3784 LNCS, pp. 224-231, 1st International Conference on ffective Computing and Intelligent Interaction, ACII 2005, Beijing, China, 05-10-22.

Vision-based recognition of hand shapes in taiwanese sign language. / Huang, Jung Ning; Hsieh, Pi Fuei; Wu, Chung-Hsien.

Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings. Vol. 3784 LNCS 2005. p. 224-231 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3784 LNCS).

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

TY - GEN

T1 - Vision-based recognition of hand shapes in taiwanese sign language

AU - Huang, Jung Ning

AU - Hsieh, Pi Fuei

AU - Wu, Chung-Hsien

PY - 2005

Y1 - 2005

N2 - The pixel-based shape representation has been sensitive to rotation. In this paper, we propose a pixel-based descriptor that is invariant with rotation and scale for the hand shape recognition in Taiwanese Sign Language (TSL). Based on the property that a hand shape is characteristic of a unique pointing direction, angle normalization is used to meet the rotation-invariant requirement. With angle normalization, the traces of class covariance matrices have been reduced almost all over the classes of hand shapes, implying a less overlap between classes. It is confirmed by the experiments that show an increase in recognition accuracy.

AB - The pixel-based shape representation has been sensitive to rotation. In this paper, we propose a pixel-based descriptor that is invariant with rotation and scale for the hand shape recognition in Taiwanese Sign Language (TSL). Based on the property that a hand shape is characteristic of a unique pointing direction, angle normalization is used to meet the rotation-invariant requirement. With angle normalization, the traces of class covariance matrices have been reduced almost all over the classes of hand shapes, implying a less overlap between classes. It is confirmed by the experiments that show an increase in recognition accuracy.

UR - http://www.scopus.com/inward/record.url?scp=33646779564&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33646779564&partnerID=8YFLogxK

M3 - Conference contribution

SN - 3540296212

SN - 9783540296218

VL - 3784 LNCS

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

SP - 224

EP - 231

BT - Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings

ER -

Huang JN, Hsieh PF, Wu C-H. Vision-based recognition of hand shapes in taiwanese sign language. In Affective Computing and Intelligent Interaction - First International Conference, ACII 2005, Proceedings. Vol. 3784 LNCS. 2005. p. 224-231. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).