Separation of multiple objects in motion images by clustering

Kohei Inoue, Kiichi Urahama

    Research output: Contribution to conferencePaperpeer-review


    A method is presented for estimating the number of rigid objects moving independently and separating the feature points tracked on a motion image into individual objects by clustering. In the method, feature points are firstly mapped into a low dimensional space suitable for grouping them into each object. In this low dimensional space, clusters are extracted sequentially by a graph spectral method. The number of clusters i.e. objects can be estimated on the basis of the variation in the cohesiveness of extracted clusters. We show by numerical experiments that the present method is robust to moderate measurement noises and distortion by perspective projection.

    Original languageEnglish
    Number of pages6
    Publication statusPublished - Jan 1 2001
    Event8th International Conference on Computer Vision - Vancouver, BC, United States
    Duration: Jul 9 2001Jul 12 2001


    Other8th International Conference on Computer Vision
    Country/TerritoryUnited States
    CityVancouver, BC

    All Science Journal Classification (ASJC) codes

    • Software
    • Computer Vision and Pattern Recognition


    Dive into the research topics of 'Separation of multiple objects in motion images by clustering'. Together they form a unique fingerprint.

    Cite this