Browse DORAS
Browse Theses
Search
Latest Additions
Creative Commons License
Except where otherwise noted, content on this site is licensed for use under a:

Generating realistic, animated human gestures in order to model, analyse and recognize Irish Sign Language

Péporté, Michèle (2009) Generating realistic, animated human gestures in order to model, analyse and recognize Irish Sign Language. Master of Science thesis, Dublin City University.

Full text available as:

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
3288Kb

Abstract

The aim of this thesis is to generate a gesture recognition system which can recognize several signs of Irish Sign Language (ISL). This project is divided into three parts. The first part provides background information on ISL. An overview of the ISL structure is a prerequisite to identifying and understanding the difficulties encountered in the development of a recognition system. The second part involves the generation of a data repository: synthetic and real-time video. Initially the synthetic data is created in a 3D animation package in order to simplify the creation of motion variations of the animated signer. The animation environment in our implementation allows for the generation of different versions of the same gesture with slight variations in the parameters of the motion. Secondly a database of ISL real-time video was created. This database contains 1400 different signs, including motion variation in each gesture. The third part details step by step my novel classification system and the associated prototype recognition system. The classification system is constructed as a decision tree to identify each sign uniquely. The recognition system is based on only one component of the classification system and has been implemented as a Hidden Markov Model (HMM).

Item Type:Thesis (Master of Science)
Date of Award:November 2009
Refereed:No
Supervisor(s):Sutherland, Alistair
Uncontrolled Keywords:gesture recognition; sign language; principal component analysis; Hidden Markov Models;
Subjects:Computer Science > Interactive computer systems
Computer Science > Visualization
Social Sciences > Gesture
Computer Science > Image processing
DCU Faculties and Centres:DCU Faculties and Schools > Faculty of Engineering and Computing > School of Computing
Use License:This item is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 3.0 License. View License
ID Code:14887
Deposited On:12 Nov 2009 10:41 by Alistair Sutherland. Last Modified 12 Nov 2009 10:41

Download statistics

Archive Staff Only: edit this record