A real-time multi face detection technique using positive-negative lines-of-face template

Yuichi Hori, Kenji Shimizu, Yutaka Nakamura, Tadahiro Kuroda

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

    23 Citations (Scopus)

    Abstract

    This paper describes a real-time multi face detection technique for color video sequences. A 3D rational skin color model and a positive-negative lines-of-face template are proposed to improve signal to noise ratio (SNR) in face detection. Steady State Genetic Algorithm (SSGA) is employed for lines-of-face detection from entire image. Hardware architecture is optimized for high-speed operation and small hardware resources. An experimental system is developed in Field Programmable Gate Array (FPGA), with only 40k gates for logic and 240k gates for memory. It detects 6 faces in real-time (30fps:every 33ms) from 320×240pixels (QVGA) color video sequences. Detection rate of 98% is achieved for 89 images including 205 faces from daily scenes.

    Original languageEnglish
    Title of host publicationProceedings of the 17th International Conference on Pattern Recognition, ICPR 2004
    EditorsJ. Kittler, M. Petrou, M. Nixon
    Pages765-768
    Number of pages4
    DOIs
    Publication statusPublished - 2004 Dec 17
    EventProceedings of the 17th International Conference on Pattern Recognition, ICPR 2004 - Cambridge, United Kingdom
    Duration: 2004 Aug 232004 Aug 26

    Publication series

    NameProceedings - International Conference on Pattern Recognition
    Volume1
    ISSN (Print)1051-4651

    Other

    OtherProceedings of the 17th International Conference on Pattern Recognition, ICPR 2004
    CountryUnited Kingdom
    CityCambridge
    Period04/8/2304/8/26

    ASJC Scopus subject areas

    • Computer Vision and Pattern Recognition

    Fingerprint Dive into the research topics of 'A real-time multi face detection technique using positive-negative lines-of-face template'. Together they form a unique fingerprint.

    Cite this