Please use this identifier to cite or link to this item:
Title: Human motion recognition
Authors: Chan, Hei Man
Department: Department of Electronic Engineering
Issue Date: 2008
Supervisor: Supervisor: Dr. Chan, K L.; Assessor: Dr. Chan, Stanley C F
Abstract: The project aims to develop a vision-based human motion recognition system specifically for recognizing the swimming styles. The captured video is pre-processed and semantic analysis is carried out to extract characteristic features from the video frames. The inference engine, which embeds high-level knowledge of the human motion, is implemented in Bayesian network (BN) and dynamic Bayesian network (DBN). BN and DBN are powerful tools for modelling recurring temporal events. Video of breaststroke and backstroke swimming styles is used to train the BN and DBN, and test the motion recognition system.
Appears in Collections:Electronic Engineering - Undergraduate Final Year Projects

Files in This Item:
File SizeFormat 
fulltext.html146 BHTMLView/Open

Items in Digital CityU Collections are protected by copyright, with all rights reserved, unless otherwise indicated.