• DocumentCode
    3211377
  • Title

    Scale and rotation invariant texture analysis based on structural property

  • Author

    Goyal, R.K. ; Goh, Wang Ling ; Mital, D.P. ; Chan, K.L.

  • Author_Institution
    Sch. of Electr. & Electron. Eng., Nanyang Technol. Inst.
  • Volume
    2
  • fYear
    1995
  • fDate
    6-10 Nov 1995
  • Firstpage
    1290
  • Abstract
    Structural properties of texture elements in a surface have successfully been applied to characterize textures. For this purpose, histograms of a texture element´s properties are used. The compactness property of an element has been used for characterization. The compactness depends upon it´s perimeter. A novel method based on perimeter contribution, to compute the value of the perimeter is presented. The perimeter value computed using the proposed method is closer to the actual, than the perimeter value computed using the conventional methods. The value of perimeter using the proposed method is also more consistent than the conventional methods. The value of compactness using the proposed method is not only closer to the actual value but also more consistent under affine transformations. The invariant histogram of compactness using the invariant value of compactness characterizes the texture invariant of scaling and rotation
  • Keywords
    computer vision; image texture; affine transformations; compactness; computer vision; invariant histogram; perimeter contribution; rotation invariant texture analysis; scale invariant texture analysis; structural property; Histograms; Humans; Image analysis; Image segmentation; Image texture analysis; Pixel; Quantization; Shape; Statistical analysis; Surface texture;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Industrial Electronics, Control, and Instrumentation, 1995., Proceedings of the 1995 IEEE IECON 21st International Conference on
  • Conference_Location
    Orlando, FL
  • Print_ISBN
    0-7803-3026-9
  • Type

    conf

  • DOI
    10.1109/IECON.1995.483983
  • Filename
    483983