• DocumentCode
    511656
  • Title

    Intelligent Restoration Algorithm to Line Similar Pixels in Noisy Image

  • Author

    Bing, Hao ; Ting-Zhi, Shen ; Hui, Tian

  • Author_Institution
    Sch. of Inf. & Electron., Beijing Inst. of Technol., Beijing, China
  • Volume
    1
  • fYear
    2009
  • fDate
    28-30 Oct. 2009
  • Firstpage
    388
  • Lastpage
    393
  • Abstract
    For restoration line similar pixels from noisy image, the disadvantages of some current image processing methods are analyzed. Then a new directional operator set with standard uniform is designed to detect the anisotropy of line pixels. Learned from manmade game schedule of washing out and sporting, a non-orthogonal and non-PR(perfect reconstruction) filtering approach is proposed. With the program of measuring, washing out, competition and tight support domain filtering, simplified by the plot of from simple to complex, straight line estimation to zigzag fitting, relative algorithm is implemented in an orderly manner. The effectiveness and consistency of proposal algorithm are checked at different noise level, and the satisfactory lines restoration results are obtained.
  • Keywords
    Radon transforms; curvelet transforms; filtering theory; image restoration; directional operator set; image processing methods; intelligent restoration algorithm; line pixels anisotropy; line similar pixels; manmade game schedule; noisy image; non-orthogonal filtering; non-perfect reconstruction filtering; relative algorithm; straight line estimation; tight support domain filtering; washing out; zigzag fitting; Algorithm design and analysis; Anisotropic magnetoresistance; Filter bank; Filtering; Image reconstruction; Image restoration; Information analysis; Noise reduction; Shape; Smart pixels; C-competition; Directional operator set; F-tight support domain smoothing filtering; M-measurement; W-wash out;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Engineering, 2009. WCSE '09. Second International Workshop on
  • Conference_Location
    Qingdao
  • Print_ISBN
    978-0-7695-3881-5
  • Type

    conf

  • DOI
    10.1109/WCSE.2009.694
  • Filename
    5403350