• DocumentCode
    296937
  • Title

    On the development of an OCR system based on curvature features

  • Author

    Gatos, B. ; Papamarkos, N.

  • Author_Institution
    Dept. of Electr. Eng., Democritus Univ. of Thrace, Xanthi, Greece
  • Volume
    1
  • fYear
    34881
  • fDate
    10-14 Jul1995
  • Firstpage
    205
  • Abstract
    This paper proposes a new method for optical character recognition which is based on curvature features. The curvature features are obtained according to the slopes of the character edge pixels. Through this technique powerful features are extracted for each character which allow us to have character recognition independent of font type and size. The proposed method can also recognize deformed fonts with high accurate results and low error rates
  • Keywords
    feature extraction; optical character recognition; OCR system; character edge pixels; classification; curvature features; deformed fonts recognition; feature extraction; low error rates; optical character recognition; Character recognition; Circuit analysis; Error analysis; Feature extraction; Filtering; Humans; Informatics; Laboratories; Office automation; Optical character recognition software;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Industrial Electronics, 1995. ISIE '95., Proceedings of the IEEE International Symposium on
  • Conference_Location
    Athens
  • Print_ISBN
    0-7803-7369-3
  • Type

    conf

  • DOI
    10.1109/ISIE.1995.496627
  • Filename
    496627