Journal of KIISE:Software and Applications (한국정보과학회논문지:소프트웨어및응용)

Korean Institute of Information Scientists and Engineers (한국정보과학회)
권호 표지

Segmentation-free Recognition of Touching Numeral Pairs

두자 접촉 숫자열의 분할 자유 인식

  • 최순만 (전북대학교 전산통계학과) ;
  • 오일석 (전북대학교 컴퓨터과학과, 영상.정보 신기술연구소)
  • Published : 2000.05.15
Download PDF
⟨ Previous Next ⟩

Abstract

Recognition of numeral fields is a very important task for many document automation applications. Conventional methods are based on the two-steps process, segmentation of touching numerals and recognition of the individual numerals. However, due to a large variation of touching types this approach has not produced a robust result. In this paper, we present a new segmentation-free method for recognizing the two touching numerals. In this approach, two touching numerals are regarded as a single pattern coming from 100 classes ('00', '01', '02', ..., '98', '99'). For the test set, we manually extract two touching numerals from the data set of NIST numeral fields. Due to the limitation of conventional neural network in case of large-set classification, we use a modular neural network and Drove its superiority through recognition experimen.

숫자열 인식은 문서 처리 자동화에서 매우 중요하다. 기존 방법들은 숫자열을 낱자 단위로 분할하는 단계와 분할된 숫자들을 인식하는 두 단계로 이루어져 있다. 그러나 이들 방법으로는 접촉 유형의 수많은 변형 때문에 만족할 만한 결과를 얻을 수 없다. 본 논문은 두자 접촉 숫자열의 분할-자유 인식 방법을 제안한다. 이 접근 방법에서는 두자 접촉 숫자열을 하나의 패턴으로 간주하여, 총 100개(‘00’, ‘01’, ‘02’, ..., ‘98’, ‘99’) 부류를 대상으로 인식한다. NIST 데이타베이스의 숫자열 필드에서 두자 접촉한 숫자열을 추출하여 실험하였다. 부류수가 방대한 경우 나타나는 기존 신경망 인식기의 한계 때문에, 모듈러 신경망을 사용하였으며 인식 실험을 통하여 우수성을 입증하였다.

Keywords

References

  1. C.Y. Suen et al., 'Computer recognition of unconstrained handwritten numerals,' Proceedings of IEEE, Vol. 80, No.7, pp. 1162-1180, 1992 https://doi.org/10.1109/5.156477
  2. H. Bunke, Advances in Structural and Syntactic Pattern Recognition, World Scientific, 1992
  3. H. Bunke, P.S.P. Wang, and H.S. Baird (Edited), Document Image Analysis, World Scientific, 1994
  4. H.S. Baird, H. Bunke, and K. Yamamoto (Edited), Structured Document Image Analysis, Springer-verlag, 1992
  5. M. Shridhar and A. Badreldin, 'Recognition of isolated and simply connected handwritten numerals,' Pattern Recognition, Vol. 19, No.1, pp. 1-12, 1986 https://doi.org/10.1016/0031-3203(86)90025-7
  6. O. Matan et al., 'Reading handwritten digits: a ZIP code recognition system,' IEEE Computers, Vol. 25, No.7, pp. 59-63, 1992 https://doi.org/10.1109/2.144441
  7. 이성환, 이응재, '순환 신경망을 이용한 연속 필기 숫자열의 통합적 분할 및 인식', 한국정보과학회 1995 추계학술발표회지, pp. 447-450, 1995
  8. H. Nishida and S. Mori, 'A model-based split-and-merge method for character string recognition,' in Document Image Analysis (Edited by Bunke et al.), World Scientific, pp. 209-226, 1994
  9. R. Fenrich, 'Segmentation of automatically located handwritten words,' Proceedings of IWFHR II, pp. 33-44, 1991
  10. J.M. Westall and M.S. Narasimha, 'Vertex directed segmentation of handwritten numerals,' Pattern Recognition, Vol. 26, No. 10, pp. 1473-1486, 1993 https://doi.org/10.1016/0031-3203(93)90153-N
  11. Z. Shi and V. Govindaraju, 'Segmentation and recognition of connected handwritten numeral strings,' Pattern Recognition, Vol. 30, No.9, pp. 1501-1505, 1997 https://doi.org/10.1016/S0031-3203(96)00118-5
  12. 최순만, 홍기천, 오일석, '무제약으로 필기한 숫자 필드의 인식에 대한 연구', 제8회 영상 처리 및 이해에 관한 워크샵 발표논문집, pp, 1-6, 1996
  13. X. Wang, V. Govindaraju, and S. Srihari, 'Holistic recognition of touching digits,' Proceedings of IWFHR VI, pp. 295-303, 1998
  14. J.T. Favata, G. Srikantau, and S.N. Srihari, 'Handwritten character/digit recognition using a multiple feature/resolution philosophy,' Proceedings of IWFHR IV, pp. 57-66, 1994
  15. C.G. Looney, Pattern Recognition using Neural Networks, Oxford University Press, 1997
  16. I.S. Oh, J.S. Lee, S.M. Choi, and K.C. Hong, 'Class-expert approach to unconstrained handwritten numeral recognition,' Proceedings of IWFHR V, pp. 95-102, 1996
  17. I.S. Oh and C.Y. Suen, 'A class-modular feedforward neural network for handwriting recognition,' IEEE Tr. on Neural Networks (submitted)
  18. N.W. Strathy, C.Y. Suen, and A. Krzyzak, 'Segmentation of handwritten digits using contour features,' Proceedings of ICDAR, pp. 577-580. 1992 https://doi.org/10.1109/ICDAR.1993.395669
  19. Z. Lu, Z. Chi, W.-C. Siu, and P. Shi, 'A Background-thinning-based approach for separating and recognizing connected handwritten digit strings,' Pattern Recognition, Vol. 32, No.6, pp. 921-933, 1999 https://doi.org/10.1016/S0031-3203(98)00123-X
  20. E. Gose, R. Johnsonbaugh and S. Jost, Pattern Recognition and Image Analysis, Prentice Hall, 1996
  21. H. Fujisawa, Y. Nakano. and K. Kurino, 'Segmentation methods for character recognition: from segmentation to document structure analysis,' Proceedings of IEEE, Vol. 80, No.7, pp. 1079-1092, 1992 https://doi.org/10.1109/5.156471
  22. M. Cheriet, Y. Huang and C. Suen, 'Background region-based algorithm for the segmentation of connected digits,' Proceedings of 11th ICPR, pp. 619-622, 1992 https://doi.org/10.1109/ICPR.1992.201853
  23. 오일석, C.Y. Suen, '광학 문자 인식을 위한 거리 특징,' 한국정보과학회논문지, Vol. 25, No.7, pp. 1028-1043, 1998
  24. A. K. Jain and B. Chandrasekaran, 'Dimensionality and sample size considerations in pattern recognition practice,' Handbook of Statistics, Vol. 2, (Edited by P.R. Krishnaiah and L.N. Kanal), North-Holland Publishing Company, pp. 835-855, 1982
  25. R. Anand, K. Mehrotra, C.K. Mohan and S. Ranka, 'An improved algorithm for neural network classification of imbalanced training sets,' IEEE Tr. on Neural Networks, Vol. 4, No.6, pp. 962-969, 1993 https://doi.org/10.1109/72.286891
  26. 오일석, 이진선, C.Y. Suen, '부류 분별에 기반한 특징 분석 및 부류-의존 특징 결합', 제10회 영상 처리 및 이해에 관한 워크샵 발표논문집, pp.169-175, 1998