Head curve matching and graffiti detection

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Vandalism is a type of the so-called low intensity crimes. However, the nuisance and damage caused by it could cost dearly to the public to clean up. It is also technically difficult in the past to detect and prosecute such a crime, for example, graffiti making. This paper presents an automatic and live graffiti detection system using CCTV cameras and innovative vision algorithms. After a brief introduction on the problem domain, the paper proceeds to proposing a system prototype through highlighting its essential operations such as human shape and head curve recognition for triggering the follow-up operations of graffiti detection. The research methodology, the devised algorithms, and the experiment designs are then explained in details. The experiments in this programme are based on real CCTV footages recorded at a bus station. The result shows that the proposed system can detect graffiti-making actions and to locate graffiti occurring areas effectively under most live illumination conditions and is a practical solution for the design of intelligent real-time CCTV surveillance systems.

Original languageEnglish
Title of host publicationBritish Machine Vision Conference, BMVC 2010 - Proceedings
PublisherBritish Machine Vision Association, BMVA
Publication statusPublished - 2010
Event21st British Machine Vision Conference 2010 - Aberystwyth, United Kingdom
Duration: 30 Aug 20102 Sep 2010
http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=5652&copyownerid=7861 (Link to Conference Details)

Conference

Conference21st British Machine Vision Conference 2010
Abbreviated title BMVC 2010
CountryUnited Kingdom
CityAberystwyth
Period30/08/102/09/10
Internet address

Fingerprint

Closed circuit television systems
Crime
Lighting
Experiments
Cameras
Costs

Cite this

Wang, J., Xu, Z., & O'Grady, M. (2010). Head curve matching and graffiti detection. In British Machine Vision Conference, BMVC 2010 - Proceedings British Machine Vision Association, BMVA.
Wang, Jing ; Xu, Zhijie ; O'Grady, Michael. / Head curve matching and graffiti detection. British Machine Vision Conference, BMVC 2010 - Proceedings. British Machine Vision Association, BMVA, 2010.
@inproceedings{75bae249dcdc476eae2206f3233cb66d,
title = "Head curve matching and graffiti detection",
abstract = "Vandalism is a type of the so-called low intensity crimes. However, the nuisance and damage caused by it could cost dearly to the public to clean up. It is also technically difficult in the past to detect and prosecute such a crime, for example, graffiti making. This paper presents an automatic and live graffiti detection system using CCTV cameras and innovative vision algorithms. After a brief introduction on the problem domain, the paper proceeds to proposing a system prototype through highlighting its essential operations such as human shape and head curve recognition for triggering the follow-up operations of graffiti detection. The research methodology, the devised algorithms, and the experiment designs are then explained in details. The experiments in this programme are based on real CCTV footages recorded at a bus station. The result shows that the proposed system can detect graffiti-making actions and to locate graffiti occurring areas effectively under most live illumination conditions and is a practical solution for the design of intelligent real-time CCTV surveillance systems.",
keywords = "Curve matching, Graffiti detection, Human recognition",
author = "Jing Wang and Zhijie Xu and Michael O'Grady",
year = "2010",
language = "English",
booktitle = "British Machine Vision Conference, BMVC 2010 - Proceedings",
publisher = "British Machine Vision Association, BMVA",

}

Wang, J, Xu, Z & O'Grady, M 2010, Head curve matching and graffiti detection. in British Machine Vision Conference, BMVC 2010 - Proceedings. British Machine Vision Association, BMVA, 21st British Machine Vision Conference 2010, Aberystwyth, United Kingdom, 30/08/10.

Head curve matching and graffiti detection. / Wang, Jing; Xu, Zhijie; O'Grady, Michael.

British Machine Vision Conference, BMVC 2010 - Proceedings. British Machine Vision Association, BMVA, 2010.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Head curve matching and graffiti detection

AU - Wang, Jing

AU - Xu, Zhijie

AU - O'Grady, Michael

PY - 2010

Y1 - 2010

N2 - Vandalism is a type of the so-called low intensity crimes. However, the nuisance and damage caused by it could cost dearly to the public to clean up. It is also technically difficult in the past to detect and prosecute such a crime, for example, graffiti making. This paper presents an automatic and live graffiti detection system using CCTV cameras and innovative vision algorithms. After a brief introduction on the problem domain, the paper proceeds to proposing a system prototype through highlighting its essential operations such as human shape and head curve recognition for triggering the follow-up operations of graffiti detection. The research methodology, the devised algorithms, and the experiment designs are then explained in details. The experiments in this programme are based on real CCTV footages recorded at a bus station. The result shows that the proposed system can detect graffiti-making actions and to locate graffiti occurring areas effectively under most live illumination conditions and is a practical solution for the design of intelligent real-time CCTV surveillance systems.

AB - Vandalism is a type of the so-called low intensity crimes. However, the nuisance and damage caused by it could cost dearly to the public to clean up. It is also technically difficult in the past to detect and prosecute such a crime, for example, graffiti making. This paper presents an automatic and live graffiti detection system using CCTV cameras and innovative vision algorithms. After a brief introduction on the problem domain, the paper proceeds to proposing a system prototype through highlighting its essential operations such as human shape and head curve recognition for triggering the follow-up operations of graffiti detection. The research methodology, the devised algorithms, and the experiment designs are then explained in details. The experiments in this programme are based on real CCTV footages recorded at a bus station. The result shows that the proposed system can detect graffiti-making actions and to locate graffiti occurring areas effectively under most live illumination conditions and is a practical solution for the design of intelligent real-time CCTV surveillance systems.

KW - Curve matching

KW - Graffiti detection

KW - Human recognition

UR - http://www.scopus.com/inward/record.url?scp=84898408964&partnerID=8YFLogxK

M3 - Conference contribution

BT - British Machine Vision Conference, BMVC 2010 - Proceedings

PB - British Machine Vision Association, BMVA

ER -

Wang J, Xu Z, O'Grady M. Head curve matching and graffiti detection. In British Machine Vision Conference, BMVC 2010 - Proceedings. British Machine Vision Association, BMVA. 2010