Sequential fuzzy cluster extraction by a graph spectral method

Kohei Inoue, Kiichi Urahama

    Research output: Contribution to journalArticlepeer-review

    34 Citations (Scopus)


    Sequential extraction method is presented for fuzzy clusters from a set of point data which is represented by a weighted graph. Nodes and links in the graph have nonnegative real weights. Memberships of data in a cluster are evaluated by the principal eigenvector of the weighted adjacency matrix of the graph. Node weights are successively reduced after extraction of clusters. The extraction of clusters is stopped when the volume of an extracted cluster reveals abrupt rebound. The method is applied to image segmentation and extraction of skin color regions from color images.

    Original languageEnglish
    Pages (from-to)699-705
    Number of pages7
    JournalPattern Recognition Letters
    Issue number7
    Publication statusPublished - Jul 1 1999

    All Science Journal Classification (ASJC) codes

    • Software
    • Signal Processing
    • Computer Vision and Pattern Recognition
    • Artificial Intelligence


    Dive into the research topics of 'Sequential fuzzy cluster extraction by a graph spectral method'. Together they form a unique fingerprint.

    Cite this