Left-object detection through background modeling

Chih Yang Lin, Chi Shiang Chan, Li Wei Kang, Kahlil Muchtar

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)

Abstract

Video surveillance systems are becoming extensively deployed in many environments due to the increasing needs of public security and crime prevention. In this paper, we propose a comprehensive solution for managing abandoned objects, which means that the system can deal with objects that are abandoned, removed, or partially occluded. The system contains two adaptive abandoned object detection (AOD) methods that are both based on the proposed texture modeling method associated with a mixture of Gaussians for a real environment. The first method is more efficient than the second one, but the latter is more robust than the former. The proposed methods have been proved to be characterized with prominent efficiency and robustness according to mathematic analyses and experimental results. The designed automatic detection system helps human operators not only to ease tedious monitoring work but also to focus only on suspicious abnormal events.

Original languageEnglish
Pages (from-to)1373-1388
Number of pages16
JournalInternational Journal of Innovative Computing, Information and Control
Volume9
Issue number4
Publication statusPublished - 2013
Externally publishedYes

Keywords

  • Background subtraction
  • Object detection
  • Video surveillance

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Information Systems
  • Computational Theory and Mathematics

Fingerprint

Dive into the research topics of 'Left-object detection through background modeling'. Together they form a unique fingerprint.

Cite this