Wavelet MACH filter for omnidirectional human action recognotion

Action recognition is important in the field of intelligent security and surveillance. However, most surveillance cameras can only capture in one direction with limited viewing angle. This research proposes an edge enhancement template based method of omnidirectional action recognition that is able...

Full description

Saved in:
Bibliographic Details
Main Author: Ang, Tyzz Kae
Format: Thesis
Published: 2012
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
id my-mmu-ep.5536
record_format uketd_dc
spelling my-mmu-ep.55362014-05-20T07:25:01Z Wavelet MACH filter for omnidirectional human action recognotion 2012-01 Ang, Tyzz Kae QA Mathematics Action recognition is important in the field of intelligent security and surveillance. However, most surveillance cameras can only capture in one direction with limited viewing angle. This research proposes an edge enhancement template based method of omnidirectional action recognition that is able to detect specific actions at a 360 degree of panoramic view. The unwarping of an omnidirectional image into a panoramic image further enables the use of existing image processing algorithm. Besides that, it also allows user to observe familiar panoramic image instead of an unfamiliar omnidirectional image. A MACH filter captures intra-class variability by synthesizing a single action MACH filter for a given action class. The proposed method, based on the wavelet MACH filter, provides additional flexibility of an adaptive choice of wavelet scale factors and, in doing so, enables the selection of the size and orientation of the smoothing function in edge enhancement to optimize the performance of the MACH filter. 2012-01 Thesis http://shdl.mmu.edu.my/5536/ http://library.mmu.edu.my/diglib/onlinedb/dig_lib.php masters Multimedia University Faculty of Engineering and Technology
institution Multimedia University
collection MMU Institutional Repository
topic QA Mathematics
spellingShingle QA Mathematics
Ang, Tyzz Kae
Wavelet MACH filter for omnidirectional human action recognotion
description Action recognition is important in the field of intelligent security and surveillance. However, most surveillance cameras can only capture in one direction with limited viewing angle. This research proposes an edge enhancement template based method of omnidirectional action recognition that is able to detect specific actions at a 360 degree of panoramic view. The unwarping of an omnidirectional image into a panoramic image further enables the use of existing image processing algorithm. Besides that, it also allows user to observe familiar panoramic image instead of an unfamiliar omnidirectional image. A MACH filter captures intra-class variability by synthesizing a single action MACH filter for a given action class. The proposed method, based on the wavelet MACH filter, provides additional flexibility of an adaptive choice of wavelet scale factors and, in doing so, enables the selection of the size and orientation of the smoothing function in edge enhancement to optimize the performance of the MACH filter.
format Thesis
qualification_level Master's degree
author Ang, Tyzz Kae
author_facet Ang, Tyzz Kae
author_sort Ang, Tyzz Kae
title Wavelet MACH filter for omnidirectional human action recognotion
title_short Wavelet MACH filter for omnidirectional human action recognotion
title_full Wavelet MACH filter for omnidirectional human action recognotion
title_fullStr Wavelet MACH filter for omnidirectional human action recognotion
title_full_unstemmed Wavelet MACH filter for omnidirectional human action recognotion
title_sort wavelet mach filter for omnidirectional human action recognotion
granting_institution Multimedia University
granting_department Faculty of Engineering and Technology
publishDate 2012
_version_ 1747829579720949760