• DocumentCode
    3488373
  • Title

    An Empirical Comparative Study of Online Handwriting Chinese Character Recognition: Simplified vs. Traditional

  • Author

    Yan Gao ; Lianwen Jin ; WeiXin Yang

  • Author_Institution
    Sch. of Electron. & Inf. Eng., South China Univ. of Technol., Guangzhou, China
  • fYear
    2013
  • fDate
    25-28 Aug. 2013
  • Firstpage
    862
  • Lastpage
    866
  • Abstract
    There are two forms of Chinese character widely used in the world, simplified Chinese and traditional Chinese. Simplified Chinese is mainly used in Chinese mainland, Singapore, Malaysia and other Southeast Asian regions, while traditional Chinese is mainly used in Hong Kong, Macao, Taiwan, Japan and so on. The simplified Chinese is transformed from traditional Chinese by simplifying the character structure and reducing the stoke number of many traditional characters, while a few remain unchanged. The simplified Chinese reduces the memory storage and makes handwriting become more convenient, while it also brings the problem that many characters are so similar that they are difficult to be recognized. In this paper, an empirical study of the handwriting simplified/traditional Chinese character recognition are carried out in order to compare the difference between these two forms of Chinese characters. The experimental results based on SCUT-Couch2009 database show that the handwriting recognition accuracy of traditional Chinese is higher than simplified Chinese, for both unchanged part and changed part. This interesting finding may bring us some cues on the issue of confusable Chinese character recognition for further study.
  • Keywords
    handwritten character recognition; natural language processing; SCUT-Couch2009 database; changed part; character structure; handwriting recognition accuracy; online handwriting Chinese character recognition; simplified Chinese character; stoke number; traditional Chinese character; unchanged part; Accuracy; Character recognition; Databases; Feature extraction; Handwriting recognition; Standards; Writing; handwriting recognition; simplified Chinese; traditional Chinese;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Document Analysis and Recognition (ICDAR), 2013 12th International Conference on
  • Conference_Location
    Washington, DC
  • ISSN
    1520-5363
  • Type

    conf

  • DOI
    10.1109/ICDAR.2013.176
  • Filename
    6628741