Image Pattern Recognition Based on Examples— A Combined Statistical and Structural-Syntactic Approach

  • Jiang Gao
  • Xiaoqing Ding
  • Jing Zheng
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1876)


An application of combined statistical and structural-syntactic approach in Chinese character recognition is presented. The algorithm adopts a structural representation for Chinese characters, but in the classification and training process, the structural matching and parameter adjustment is conducted in a statistical way. Different from the conventional structural approaches, in this system, only a few predefined “knowledge” is required. In most cases, knowledge acquisition is simplified to “memorization” of examples, and the parameters for classification can be refined using statistical training. In this way it avoids the main difficulties inherent in the implementation of classification systems based on structural features. Compared with conventional statistical algorithms, the algorithm is based on a structural model of image patterns, so it has approximately all the advantages of structural pattern recognition algorithms. A prototype system has been realized based on this strategy, and the effectiveness of the method is verified by the experimental results.


  1. 1.
    Hildebrandt, T.H. and W. Liu, Optical recognition of handwritten Chinese characters: advances since 1980. Pattern Recognition, Vol. 26, No. 2, pp. 205–225, 1993.CrossRefGoogle Scholar
  2. 2.
    F. Kimura, et al., Improvement of Handwritten Japanese Character Recognition Using Weighted Direction Code Histogram, Pattern Recognition, Vo1. 30, No. 8, pp. 1329–1337, 1997.CrossRefGoogle Scholar
  3. 3.
    Rocha, J. and T. Pavlidis, A shape analysis model with applications to a character recognition system. IEEE Trans. PAMI, Vol. 16, pp. 393–404, 1994.Google Scholar
  4. 4.
    Hong Dong, Xiaoqing Ding and Youshou Wu, The ARG description of Chinese characters and a radical extracting method based on this description. The Chinese Information Journal, Vol. 1, No. 3, pp. 47–52, 1987.Google Scholar
  5. 5.
    C.C Tappert, The state of the art in on-line handwriting recognition, IEEE Trans. PAMI, Vol. 12, No. 8, pp. 787–808, 1990.Google Scholar
  6. 6.
    F. Nouboud and R. Plamondon, On-line recognition of hand-printed characters: survey and beta tests, Pattern Recognition, Vol. 23, No. 9, pp. 1031–1044, 1990.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Jiang Gao
    • 1
  • Xiaoqing Ding
    • 1
  • Jing Zheng
    • 2
  1. 1.Institute of Image and Graphics, Department of Electronic EngineeringTsinghua UniversityBeijingPRC
  2. 2.Speech Technology and Research LabSRI InternationalMenlo Park

Personalised recommendations