school

UM E-Theses Collection (澳門大學電子學位論文庫)

check Full Text
Title

Real-time hand gesture recognition using motion tracking

English Abstract

This paper introduces an attractive alternative human computer interaction interface in future applications, which is the gesture performed by the user. Especially we focus on hand gestures by using its motion representation and several techniques and related works in recent years are reviewed. A gesture recognition system of 10 hand signed digits is proposed in our research, in which we firstly compared different solutions of detecting active hand in a video frame and concluded our skin-subtraction approach, and then locations of hand in a sequence of frames are tracked to extract the feature of motion track. Finally we used the histogram distribution models to recognize each location track as one of ten digits. Our proposed system achieves a recognition rate of 97.33% and also supports for the real-time application.

Issue date

2010

Author

Zhu, Hong Min

Faculty

Faculty of Science and Technology

Degree

M.Sc.

Subject

Optical pattern recognition

Image processing -- Digital techniques

Gesture

Real-time data processing

Software Engineering -- Department of Computer and Information Science

Supervisor

Pun, Chi Man

Files In This Item

Full-text (Intranet)

Full-text (internet)

Location
1/F Zone C
Library URL
991005548919706306