Advertisement

Fuzzy Sets pp 195-221 | Cite as

Experiment on Character Recognition Using Fuzzy Filters

  • Paul P. Wang
  • C. Y. Wang

Abstract

The theory of fuzzy set first introduced by Lofti Zadeh is employed in designing filters. The filters are used in the automatic recognition of printed alphanumeric character patterns by comparing a dictionary of 37 standard masks to the noisy data pattern and selecting the mask which fits best. The basic building blocks of our proposed fuzzy filters consists of Boolean and fuzzy logic operations, Boolean and fuzzy relational operations, and mixed functions of fuzzy variables. Extensive computer simulation was performed to evaluate the performance. The results are very encouraging for both versions of our design; (i) nearly perfect recognition rate was achieved with up to 22.86% of picture cells corrupted by noise, and (ii) better than 90% correct recognition rate was observed with up to 31.43% (Filter I) and 37.14% (Filter II) of picture cells contaminated by noise. We believe that fuzzy set theory will be a very powerful tool in filter design, providing simplicity in implementation of both hardware and software.

Keywords

Membership Function Character Recognition Fuzzy Variable Fuzzy Graph Alphanumeric Character 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    M. M. Gupta, G. N. Saridis, and B. R. Gaines, “Fuzzy Automata and Decision Process,” American Elsevier, North-Holland, Inc., New York (1977).Google Scholar
  2. 2.
    A. Kaufmann, “Theory of Fuzzy Subsets,” Academic Press, New York (1975).Google Scholar
  3. 3.
    R. E. Bellman, R. Kalaba, and L. A. Zadeh, Abstraction and Pattern Classification, J. of Math. Anal, and Applications, Vol. 13, pp. 1–7 (1966).CrossRefGoogle Scholar
  4. 4.
    AN-85 Saving ROM’s in High Resolution Dot-Matrix Displays and Printers, “National Semiconductor Memory Products Databook,” pp. 11, 37–38.Google Scholar
  5. 5.
    P. Siy and C. C. Chen, Fuzzy Logic for Handwritten Numeral Character Recognition,’ IEEE Transactions on Systems, Man and Cybernetics, pp. 570–575 (Nov. 1974).Google Scholar
  6. 6.
    W. J. M. Kickert and H. Koppelaar, Application of Fuzzy Set Theory to Syntactic Pattern Recognition of Handwritten Capitals, IEEE Trans, on Sys., Man and Cybernetics, pp. 148–151 (Feb. 1976).Google Scholar
  7. 7.
    E. Hisdal, Conditional Possibilities, Independence and Noninteraction, in: “Fuzzy Sets and Systems,” Vol. 1, pp. 283–397, North-Holland Publishing Company, New York (1978).Google Scholar
  8. 8.
    S. Tamura, S. Higuchi, and K. Tanaka, Pattern Classification Based on Fuzzy Relations, IEEE Trans, on Sys., Man and Cybernetics, Vol. SMC-1, pp. 61–66 (Jan. 1971).CrossRefGoogle Scholar
  9. 9.
    M. G. Thomasson, Finite Fuzzy Automata, Regular Fuzzy Languages and Pattern Recognition, J. of Pattern Recognition, Vol. 5, pp. 383–390 (1973).CrossRefGoogle Scholar
  10. 10.
    J. G. Dunn, A Graph Theoretic Analysis of Pattern Classification via Tamura’s Fuzzy Relation, IEEE Trans, on Sys., Man and Cybernetics, Vol. SMC-4, pp. 310–313 (1974).Google Scholar
  11. 11.
    A. Papoulis, “Probability, Random Variables and Stochastic Processes,” McGraw-Hill Book Co., New York (1965).Google Scholar

Copyright information

© Plenum Press, New York 1980

Authors and Affiliations

  • Paul P. Wang
    • 1
  • C. Y. Wang
    • 1
  1. 1.Department of Electrical EngineeringDuke UniversityDurhamUSA

Personalised recommendations