• DocumentCode
    580317
  • Title

    Anisotropic diffusion filter without conductivity parameters

  • Author

    Demirci, Recep ; Guvenc, Ugur ; Tanyeri, Ufuk

  • Author_Institution
    Comput. Eng. Dept., Gazi Univ., Ankara, Turkey
  • fYear
    2012
  • fDate
    12-14 Oct. 2012
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    Two novel automatic anisotropic diffusion filters without conduction parameters were proposed in this paper. In the proposed first method, initially, the gradient of the noisy image to be filtered was calculated by Sobel operator and then the center of gravity of gradient image was used as diffusivity parameter, K and edge threshold. It was observed that the center of gravity of gradient image changed with noise. In the second approach, the complement of gradient image was directly employed as heat conductivity parameter in image filter. Consequently, the user dependency was avoided. The proposed image filter was tested with well-known test images.
  • Keywords
    edge detection; filtering theory; gradient methods; image denoising; image segmentation; Sobel operator; automatic anisotropic diffusion filters; center of gravity; diffusivity parameter; edge threshold; heat conductivity parameter; noisy image gradient filtering; Anisotropic magnetoresistance; Conductivity; Image edge detection; Logistics; Materials; Maximum likelihood detection; Nonlinear filters;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    System Theory, Control and Computing (ICSTCC), 2012 16th International Conference on
  • Conference_Location
    Sinaia
  • Print_ISBN
    978-1-4673-4534-7
  • Type

    conf

  • Filename
    6379265