English  |  正體中文  |  简体中文  |  Items with full text/Total items : 51931/87076 (60%)
Visitors : 8492573      Online Users : 320
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/108295


    Title: A fast face detection method for illumination variant condition
    Authors: C.-H. Hsia;J.-S. Chiang;C.-Y. Lin
    Keywords: Illumination variant face detection;Adaboost;Neural network;Modi_ed census transform;Real-time detection
    Date: 2015/12/01
    Issue Date: 2016-11-22 02:10:45 (UTC+8)
    Publisher: Sharif University of Technology
    Abstract: General boosting algorithms for face detection use rectangular features. To obtain a better performance, it needs more training samples and may generate an unpredictable number of features. Besides using pixel values, which are easily affected by illumination, to calculate the rectangular features, it usually needs to preprocess the data before calculating the values of the features. Such an approach may increase computation time. To overcome the drawbacks, we propose a new solution based on the Adaboost algorithm and the Back Propagation Network (BPN) of a Neural Network (NN), combining local and global features with cascade architecture to detect human faces. We use the Modified Census Transform (MCT) feature, which belongs to texture features and is less sensitive to illumination, for local feature calculation. In this approach, it is not necessary to preprocess each sub-window of the image. For classification, we use the structure of the hierarchical feature to control the number of features. With only MCT, it is easy to misjudge faces and, therefore, in this work, we include the brightness information of global features to eliminate the False Positive (FP) regions. As a result, the proposed approach can have a Detection Rate (DR) of 99%, an FPs of only 11, and detection speed of 27.92 Frames Per Second (FPS).
    Relation: Scientia Iranica B 22(6), pp.2081-2091
    DOI: 
    Appears in Collections:[電機工程學系暨研究所] 期刊論文

    Files in This Item:

    File Description SizeFormat
    index.html0KbHTML85View/Open

    All items in 機構典藏 are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback