Title

A vision-based approach for human hand tracking and gesture recognition.

Date of Award

2004

Degree Type

Thesis

Degree Name

M.Sc.

Department

Computer Science

First Advisor

Yuan, Xiaobu,

Keywords

Computer Science.

Rights

CC BY-NC-ND 4.0

Abstract

Hand gesture interface has been becoming an active topic of human-computer interaction (HCI). The utilization of hand gestures in human-computer interface enables human operators to interact with computer environments in a natural and intuitive manner. In particular, bare hand interpretation technique frees users from cumbersome, but typically required devices in communication with computers, thus offering the ease and naturalness in HCI. Meanwhile, virtual assembly (VA) applies virtual reality (VR) techniques in mechanical assembly. It constructs computer tools to help product engineers planning, evaluating, optimizing, and verifying the assembly of mechanical systems without the need of physical objects. However, traditional devices such as keyboards and mice are no longer adequate due to their inefficiency in handling three-dimensional (3D) tasks. Special VR devices, such as data gloves, have been mandatory in VA. This thesis proposes a novel gesture-based interface for the application of VA. It develops a hybrid approach to incorporate an appearance-based hand localization technique with a skin tone filter in support of gesture recognition and hand tracking in the 3D space. With this interface, bare hands become a convenient substitution of special VR devices. Experiment results demonstrate the flexibility and robustness introduced by the proposed method to HCI.Dept. of Computer Science. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis2004 .L8. Source: Masters Abstracts International, Volume: 43-03, page: 0883. Adviser: Xiaobu Yuan. Thesis (M.Sc.)--University of Windsor (Canada), 2004.