• DocumentCode
    2030692
  • Title

    PATRAM - a handwritten word processor for Indian languages

  • Author

    Madduri, Kamesh ; Aparna, K.H. ; Chakravarthy, V.S.

  • Author_Institution
    Indian Inst. of Technol., Madras, India
  • fYear
    2004
  • fDate
    26-29 Oct. 2004
  • Firstpage
    557
  • Lastpage
    562
  • Abstract
    In this paper, we present PATRAM, a comprehensive handwritten wordprocessing system specially designed for Indian languages. In typical existing ink capture programs, handwriting is entered in a ´paint-on-a-wall´ style without any syntactical structure imposed on the handwritten text. The PATRAM document page is classified into text (ruled space) and drawing (unruled space) areas. While inking the text region, the writer stays within the lines. The drawing region does not have such a restriction. While this arrangement accommodates unconstrained inking (in drawing area), it provides a more intuitive and convenient editing style for the text area. The standard operations including insert, cut, copy, paste, delete, select, justify etc. are defined which operate on text or drawing objects in characteristic fashion. The document is saved in a specific format (.pat) which makes off-line character recognition possible. We finally demonstrate recognition capability in Tamil language.
  • Keywords
    handwritten character recognition; natural languages; word processing; Indian languages; PATRAM document page; Tamil language; handwritten text; handwritten word processor; ink capture programs; offline character recognition; paint-on-a-wall style; Character recognition; Conferences; Handwriting recognition; Ink; Keyboards; Natural languages; Personal communication networks; Personal digital assistants; Text processing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Frontiers in Handwriting Recognition, 2004. IWFHR-9 2004. Ninth International Workshop on
  • ISSN
    1550-5235
  • Print_ISBN
    0-7695-2187-8
  • Type

    conf

  • DOI
    10.1109/IWFHR.2004.82
  • Filename
    1363970