• DocumentCode
    529405
  • Title

    An image processing for surface inspection

  • Author

    Fujidai, Tomoko ; Yamaguchi, Jun´ichi ; Nagai, Motoki

  • Author_Institution
    Kagawa Univ., Takamatsu, Japan
  • fYear
    2010
  • fDate
    18-21 Aug. 2010
  • Firstpage
    3601
  • Lastpage
    3604
  • Abstract
    In this paper, the authors describe an image processing method for identification of an erosion on concrete structure surface. Many erosion snapshots by hand are handled. In this case, it is difficult to verify the erosion image because of camera shooting condition change (angle, distance, brightness, etc.). This method is made up of two steps. First step is shape verification using Matched filter. Second step is identification of Hue-Saturation distribution of the candidate erosion detected by first step. As a result of an experiment, it was confirmed that the method functioned usefully for identification of the erosion image in snapshot whose shooting condition is unknown.
  • Keywords
    cameras; colour graphics; concrete; erosion; feature extraction; filtering theory; image processing; matched filters; shape recognition; camera shooting condition; concrete structure surface erosion; erosion image; erosion snapshot; hue saturation distribution; image processing; matched filter; shape verification; surface inspection; Cameras; Concrete; Correlation; Image color analysis; Matched filters; Shape; Surface treatment; Camera shooting condition; HSV color system; Image verification (key words); Matched filter; Perspective Transformation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    SICE Annual Conference 2010, Proceedings of
  • Conference_Location
    Taipei
  • Print_ISBN
    978-1-4244-7642-8
  • Type

    conf

  • Filename
    5602651