Regularized Online Mixture of Gaussians for Background Subtraction

Hongbin Wang, Paul Miller

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

16 Citations (Scopus)
456 Downloads (Pure)

Abstract

Mixture of Gaussians (MoG) modelling [13] is a popular approach to background subtraction in video sequences. Although the algorithm shows good empirical performance, it lacks theoretical justification. In this paper, we give a justification for it from an online stochastic expectation maximization (EM) viewpoint and extend it to a general framework of regularized online classification EM for MoG with guaranteed convergence. By choosing a special regularization function, l1 norm, we derived a new set of updating equations for l1 regularized online MoG. It is shown empirically that l1 regularized online MoG converge faster than the original online MoG .
Original languageEnglish
Title of host publicationAdvanced Video and Signal-Based Surveillance (AVSS), 2011 8th IEEE International Conference on
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages249-254
Number of pages6
ISBN (Print)978-1-4577-0845-9
DOIs
Publication statusPublished - Sep 2011
Event8th IEEE International Conference on Advanced Video and Signal-Based Surveillance - Klagenfurt, Austria
Duration: 30 Aug 201102 Sep 2011

Conference

Conference8th IEEE International Conference on Advanced Video and Signal-Based Surveillance
CountryAustria
CityKlagenfurt
Period30/08/201102/09/2011

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Fingerprint Dive into the research topics of 'Regularized Online Mixture of Gaussians for Background Subtraction'. Together they form a unique fingerprint.

Cite this