Snatch theft detection using low level features

Norazlin Ibrahim, Siti Salasiah Mokri, Lee Yee Siong, Mohd. Marzuki Mustafa, Aini Hussain

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

6 Citations (Scopus)

Abstract

In practice it is difficult to diagnose events based on the ability to segment the individual persons in a crowd.. The used of low level features is seemed to be more effective to identify the abnormality situation. This paper presents the detection of snatch theft in pedestrian crowd movement. It is based on the features extracted from the computation of optical flow for sequence of video frame. Kalman filter is used to detect the start and the end of of possible snatching events. The event is classified based on the distribution of the optical flow vectors before and after the events using vector matching and SVM classification. The algorithm has been tested on simulated events, and showed a good detection rate of snatch theft events..

Original languageEnglish
Title of host publicationWCE 2010 - World Congress on Engineering 2010
Pages862-866
Number of pages5
Volume2
Publication statusPublished - 2010
EventWorld Congress on Engineering 2010, WCE 2010 - London
Duration: 30 Jun 20102 Jul 2010

Other

OtherWorld Congress on Engineering 2010, WCE 2010
CityLondon
Period30/6/102/7/10

Fingerprint

Optical flows
Kalman filters

Keywords

  • Kalman filter
  • Optical flow
  • Snatch theft
  • SVM

ASJC Scopus subject areas

  • Computer Science(all)
  • Engineering(all)

Cite this

Ibrahim, N., Mokri, S. S., Siong, L. Y., Mustafa, M. M., & Hussain, A. (2010). Snatch theft detection using low level features. In WCE 2010 - World Congress on Engineering 2010 (Vol. 2, pp. 862-866)

Snatch theft detection using low level features. / Ibrahim, Norazlin; Mokri, Siti Salasiah; Siong, Lee Yee; Mustafa, Mohd. Marzuki; Hussain, Aini.

WCE 2010 - World Congress on Engineering 2010. Vol. 2 2010. p. 862-866.

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

Ibrahim, N, Mokri, SS, Siong, LY, Mustafa, MM & Hussain, A 2010, Snatch theft detection using low level features. in WCE 2010 - World Congress on Engineering 2010. vol. 2, pp. 862-866, World Congress on Engineering 2010, WCE 2010, London, 30/6/10.
Ibrahim N, Mokri SS, Siong LY, Mustafa MM, Hussain A. Snatch theft detection using low level features. In WCE 2010 - World Congress on Engineering 2010. Vol. 2. 2010. p. 862-866
Ibrahim, Norazlin ; Mokri, Siti Salasiah ; Siong, Lee Yee ; Mustafa, Mohd. Marzuki ; Hussain, Aini. / Snatch theft detection using low level features. WCE 2010 - World Congress on Engineering 2010. Vol. 2 2010. pp. 862-866
@inproceedings{3189036c1e9740848c130ce5b9ea50ae,
title = "Snatch theft detection using low level features",
abstract = "In practice it is difficult to diagnose events based on the ability to segment the individual persons in a crowd.. The used of low level features is seemed to be more effective to identify the abnormality situation. This paper presents the detection of snatch theft in pedestrian crowd movement. It is based on the features extracted from the computation of optical flow for sequence of video frame. Kalman filter is used to detect the start and the end of of possible snatching events. The event is classified based on the distribution of the optical flow vectors before and after the events using vector matching and SVM classification. The algorithm has been tested on simulated events, and showed a good detection rate of snatch theft events..",
keywords = "Kalman filter, Optical flow, Snatch theft, SVM",
author = "Norazlin Ibrahim and Mokri, {Siti Salasiah} and Siong, {Lee Yee} and Mustafa, {Mohd. Marzuki} and Aini Hussain",
year = "2010",
language = "English",
isbn = "9789881821072",
volume = "2",
pages = "862--866",
booktitle = "WCE 2010 - World Congress on Engineering 2010",

}

TY - GEN

T1 - Snatch theft detection using low level features

AU - Ibrahim, Norazlin

AU - Mokri, Siti Salasiah

AU - Siong, Lee Yee

AU - Mustafa, Mohd. Marzuki

AU - Hussain, Aini

PY - 2010

Y1 - 2010

N2 - In practice it is difficult to diagnose events based on the ability to segment the individual persons in a crowd.. The used of low level features is seemed to be more effective to identify the abnormality situation. This paper presents the detection of snatch theft in pedestrian crowd movement. It is based on the features extracted from the computation of optical flow for sequence of video frame. Kalman filter is used to detect the start and the end of of possible snatching events. The event is classified based on the distribution of the optical flow vectors before and after the events using vector matching and SVM classification. The algorithm has been tested on simulated events, and showed a good detection rate of snatch theft events..

AB - In practice it is difficult to diagnose events based on the ability to segment the individual persons in a crowd.. The used of low level features is seemed to be more effective to identify the abnormality situation. This paper presents the detection of snatch theft in pedestrian crowd movement. It is based on the features extracted from the computation of optical flow for sequence of video frame. Kalman filter is used to detect the start and the end of of possible snatching events. The event is classified based on the distribution of the optical flow vectors before and after the events using vector matching and SVM classification. The algorithm has been tested on simulated events, and showed a good detection rate of snatch theft events..

KW - Kalman filter

KW - Optical flow

KW - Snatch theft

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=79959846629&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79959846629&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9789881821072

VL - 2

SP - 862

EP - 866

BT - WCE 2010 - World Congress on Engineering 2010

ER -