South African sign language recognition using feature vectors and Hidden Markov Models

South African sign language recognition using feature vectors and Hidden Markov Models

Masters of Science === This thesis presents a system for performing whole gesture recognition for South African Sign Language. The system uses feature vectors combined with Hidden Markov models. In order to constuct a feature vector, dynamic segmentation must occur to extract the signer's hand...

Full description

Bibliographic Details
Main Author: Naidoo, Nathan Lyle
Other Authors: Connan, James
Language:en
Published: University of the Western Cape 2013
Subjects:
Optical pattern recognition
Mathematical models
Image processing
Digital techniques
Markov processes
Online Access:http://hdl.handle.net/11394/2527
id ndltd-netd.ac.za-oai-union.ndltd.org-uwc-oai-etd.uwc.ac.za-11394-2527
record_format oai_dc
spelling ndltd-netd.ac.za-oai-union.ndltd.org-uwc-oai-etd.uwc.ac.za-11394-25272017-08-02T04:00:13Z South African sign language recognition using feature vectors and Hidden Markov Models Naidoo, Nathan Lyle Connan, James Dept. of Computer Science Faculty of Science Optical pattern recognition Mathematical models Image processing Digital techniques Markov processes Masters of Science This thesis presents a system for performing whole gesture recognition for South African Sign Language. The system uses feature vectors combined with Hidden Markov models. In order to constuct a feature vector, dynamic segmentation must occur to extract the signer's hand movements. Techniques and methods for normalising variations that occur when recording a signer performing a gesture, are investigated. The system has a classification rate of 69%. South Africa 2013-12-10T14:10:46Z 2011/02/17 08:20 2011/02/17 2013-12-10T14:10:46Z 2010 Thesis http://hdl.handle.net/11394/2527 en University of the Western Cape University of the Western Cape
collection NDLTD
language en
sources NDLTD
topic Optical pattern recognition
Mathematical models
Image processing
Digital techniques
Markov processes
spellingShingle Optical pattern recognition
Mathematical models
Image processing
Digital techniques
Markov processes
Naidoo, Nathan Lyle
South African sign language recognition using feature vectors and Hidden Markov Models
description Masters of Science === This thesis presents a system for performing whole gesture recognition for South African Sign Language. The system uses feature vectors combined with Hidden Markov models. In order to constuct a feature vector, dynamic segmentation must occur to extract the signer's hand movements. Techniques and methods for normalising variations that occur when recording a signer performing a gesture, are investigated. The system has a classification rate of 69%. === South Africa
author2 Connan, James
author_facet Connan, James
Naidoo, Nathan Lyle
author Naidoo, Nathan Lyle
author_sort Naidoo, Nathan Lyle
title South African sign language recognition using feature vectors and Hidden Markov Models
title_short South African sign language recognition using feature vectors and Hidden Markov Models
title_full South African sign language recognition using feature vectors and Hidden Markov Models
title_fullStr South African sign language recognition using feature vectors and Hidden Markov Models
title_full_unstemmed South African sign language recognition using feature vectors and Hidden Markov Models
title_sort south african sign language recognition using feature vectors and hidden markov models
publisher University of the Western Cape
publishDate 2013
url http://hdl.handle.net/11394/2527
work_keys_str_mv AT naidoonathanlyle southafricansignlanguagerecognitionusingfeaturevectorsandhiddenmarkovmodels
_version_ 1718510386634817536

Similar Items