• DocumentCode
    3403552
  • Title

    A Two-dimensional Structure Analysis Method of Printing Mathematical Formula Based On Characteristic Characters

  • Author

    Li, Yonghua ; Wang, Kejun ; Tang, Liqun ; Wu, Junfei

  • Author_Institution
    Harbin Eng. Univ., Harbin
  • fYear
    2007
  • fDate
    5-8 Aug. 2007
  • Firstpage
    953
  • Lastpage
    957
  • Abstract
    The biggest difficulty in mathematical formula recognition is two-dimensional structure analysis. A two-dimensional structure analysis method of printing mathematical formula based on characteristic character is presented in the paper. This method presents the notion of character and divides the characters that are always been seen in the mathematical formula into five kinds, it also presents a block incision method to recognize printing formula based on character. The paper designs different kinds of recognition algorithms based on different characters to complete the two-dimensional structure analysis. By means of the contrastive test of three hundred printing mathematical formulas which have different complexity of superscript and subscript field, the results indicates that the method in this paper gets the highest overall recognition rate and lower mistake rate in the false recognition test.
  • Keywords
    character recognition; mathematics computing; text analysis; 2D structure analysis; block incision method; characteristic characters; contrastive test; mathematical formula recognition; printing formula recognition; subscript; superscript; Algorithm design and analysis; Automation; Character recognition; Educational institutions; Information analysis; Mechatronics; Optical character recognition software; Printing; Testing; Writing; characteristic character; contrastive test; two-dimensional structure analysis;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Mechatronics and Automation, 2007. ICMA 2007. International Conference on
  • Conference_Location
    Harbin
  • Print_ISBN
    978-1-4244-0828-3
  • Electronic_ISBN
    978-1-4244-0828-3
  • Type

    conf

  • DOI
    10.1109/ICMA.2007.4303675
  • Filename
    4303675