Beyond static detectors: a Bayesian approach to fusing long-term motion with appearance for robust people detection in highly cluttered scenes

Jianguo Zhang, Shaogang Gong

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

    Abstract

    In this work we present a framework for robust people de-tection in highly cluttered scenes with low resolution im-age sequences. Our model utilises both human appearance and their long-term motion information through a fusion formulated in a Bayesian framework. In particular, peo-ple appearance is modeled by histograms of oriented gra-dients. Motion information is computed via an improved background modeling by spatial motion constrains. Exper-iments demonstrate that our method reduces significantly the false positive rate compared to that of a state of the art human detector under very challenging conditions.
    Original languageEnglish
    Title of host publicationProceedings of the Sixth International IEEE Workshop on Visual Surveillance
    EditorsPascal Fua, Stephen Maybank, Graeme A. Jones
    Place of PublicationKingston-Upon-Thames
    PublisherIEEE
    Pages121-128
    Number of pages8
    ISBN (Print)9780955300301, 0955300304
    Publication statusPublished - 2006
    Event6th IEEE International Workshop on Visual Surveillance - Graz, Austria
    Duration: 13 May 200613 May 2006

    Conference

    Conference6th IEEE International Workshop on Visual Surveillance
    CountryAustria
    CityGraz
    Period13/05/0613/05/06

    Fingerprint Dive into the research topics of 'Beyond static detectors: a Bayesian approach to fusing long-term motion with appearance for robust people detection in highly cluttered scenes'. Together they form a unique fingerprint.

    Cite this