Please use this identifier to cite or link to this item:
Title: Development of human activity recognition system
Authors: Lim, Aaron Philip Jian Hui
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2017
Abstract: Human activity recognition is an important area of research in today’s complex world as it can be applied to many areas, and play a significant role in human-human, and human-computer relationship. Being able to recognize another person’s activity is one of the main area of focus for computer vision and machine learning. As such, many applications such as human-computer interactions, video surveillance systems, and robotics require an activity recognition system . The aim of human activity recognition is to recognize and identify human activities in real time settings. To be able to identify and recognize human activity is however often difficult and challenging as human activity is highly complex and diverse . Common place activities such as walking might be easy to identify, however, more complex actions such as playing a sport of doing a specific martial art move is more difficult to recognize, and might be decomposed into smaller actions to make it easier to identify. To be able to fully develop an automated system to be able to recognize human activity with a low margin of error is challenging due to issues such as background noises, changes in scale, video resolution, lighting and appearance etc. Therefore, most studies prefer the use of an isolated area with a clear background where a person can perform an action freely. There are several methods used to build such human activity recognition models such as the Motion History Image (MHI), coupled with the Local Binary Pattern (LBP).
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
FYP Report.pdf
  Restricted Access
1.17 MBAdobe PDFView/Open

Page view(s)

Updated on May 12, 2021

Download(s) 50

Updated on May 12, 2021

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.