• DocumentCode
    2411579
  • Title

    On the discretization of parameter domain in Hough transformation

  • Author

    Zhang, Ming

  • Author_Institution
    Dept. of Radiol., Chicago Univ., IL, USA
  • Volume
    2
  • fYear
    1996
  • fDate
    25-29 Aug 1996
  • Firstpage
    527
  • Abstract
    Although the basic principle of Hough transformation can be described accurately in continuous spaces, its application is often conducted in digitized ones. The discretization of both a spatial image and the related transformation parameters will result in positioning errors in the parameter domain that affect the accumulation through which the Hough transformation functions. This makes the discretization of its parameter domain an important issue. Its resolution needs to be carefully selected to assure the effective concentration of the accumulation. In this paper, the effects of the digitization of both the spatial and the parameter domain on the resolution of the latter are analyzed numerically, and a procedure to determine the resolution, and therefore the discretization, of the parameter domain is proposed. Computer simulations show the effect of this consideration, and its importance is also demonstrated by applying successfully the Hough transformation to texture analysis in mammographic image processing with the appropriate digitization of its parameter domain
  • Keywords
    Hough transforms; computer vision; diagnostic radiography; image texture; medical image processing; Hough transform; discretization; image texture analysis; mammographic images; parameter domain; spatial image processing; Computer simulation; Degradation; Equations; Image analysis; Image processing; Image texture analysis; Laboratories; Motion analysis; Shape; Spatial resolution;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Pattern Recognition, 1996., Proceedings of the 13th International Conference on
  • Conference_Location
    Vienna
  • ISSN
    1051-4651
  • Print_ISBN
    0-8186-7282-X
  • Type

    conf

  • DOI
    10.1109/ICPR.1996.546880
  • Filename
    546880