• DocumentCode
    2863736
  • Title

    Photomask Defect Extraction by Using Difference between a Reference Image and a Test Image after Illumination Adjustment

  • Author

    Youngmin Ha

  • Author_Institution
    POSTECH, Pohang
  • fYear
    2007
  • fDate
    11-13 Oct. 2007
  • Firstpage
    242
  • Lastpage
    248
  • Abstract
    A new method for extracting a photomask defect by using two optical photomask images, a reference image and a test image, is described in this paper. If the same scale, the same height-to-width ratio, zero parallel translation, and zero rotation between two images are assumed, then a large pixel value of the absolute difference image between the images would indicate the existence of a defect at the pixel. However, not only a true defect but a false defect also causes some large pixel values in the absolute difference image due to illumination difference between the reference image and the test image. To adjust the illumination of the reference image to that of the test image, a dynamic programming is used, and the defect in the test image is segmented.
  • Keywords
    dynamic programming; flaw detection; image matching; image segmentation; masks; dynamic programming; illumination adjustment; image segmentation; photomask defect extraction; photomask image difference; Data mining; Dynamic programming; Image segmentation; Inspection; Lighting; Pervasive computing; Pixel; Shape; Silicon; System testing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Intelligent Pervasive Computing, 2007. IPC. The 2007 International Conference on
  • Conference_Location
    Jeju City
  • Print_ISBN
    978-0-7695-3006-2
  • Type

    conf

  • DOI
    10.1109/IPC.2007.53
  • Filename
    4438433