DOI QR코드

DOI QR Code

Online Handwritten Digit Recognition by Smith-Waterman Alignment

Smith-Waterman 정렬 알고리즘을 이용한 온라인 필기체 숫자인식

  • Mun, Won-Ho (Dept. of Computer Engineering, Pusan National University) ;
  • Choi, Yeon-Seok (Dept. of Computer Engineering, Pusan National University) ;
  • Lee, Sang-Geol (Dept. of Computer Engineering, Pusan National University) ;
  • Cha, Eui-Young (Dept. of Computer Engineering, Pusan National University)
  • 문원호 (부산대학교 컴퓨터공학과) ;
  • 최연석 (부산대학교 컴퓨터공학과) ;
  • 이상걸 (부산대학교 컴퓨터공학과) ;
  • 차의영 (부산대학교 컴퓨터공학과)
  • Received : 2011.04.29
  • Accepted : 2011.06.21
  • Published : 2011.09.30

Abstract

In this paper, we propose an efficient on-line handwritten digit recognition base on Convex-Concave curves feature which is extracted by a chain code sequence using Smith-Waterman alignment algorithm. The time sequential signal from mouse movement on the writing pad is described as a sequence of consecutive points on the x-y plane. So, we can create data-set which are successive and time-sequential pixel position data by preprocessing. Data preprocessed is used for Convex-Concave curves feature extraction. This feature is scale-, translation-, and rotation-invariant. The extracted specific feature is fed to a Smith-Waterman alignment algorithm, which in turn classifies it as one of the nine digits. In comparison with backpropagation neural network, Smith-Waterman alignment has the more outstanding performance.

본 논문에서는 필기체 문자의 Convex-Concave한 곡선 특징을 문자로 변환하고 추출된 문자를 Smith-Waterman 정렬 알고리즘을 이용하여 온라인 필기체 숫자 인식 방법을 제안한다. 필기체 숫자 인식을 위한 입력 데이터는 시간에 순서적인 좌표로 순서화하고 전처리의 입력데이터로 적용된다. 필기자의 개성이 표현된 필기체 문자는 크기, 회전, 곡선 비율이 다양한 형태로 나타난다. 따라서 본 논문에서는 곡선의 Convex-Concave 특징을 이용하여 크기, 회전에 강인한 특징을 추출한다. 추출된 특징은 문자로 변환하고 Smith-Waterman 정렬 알고리즘의 입력데이터로 적용한다. 본 논문에서는 실시간 필기체 숫자를 대상으로 실험한 결과, 오류역전파 신경 회로망을 적용한 것과 비교하여 제안된 방법이 좋은 성능을 보였다.

Keywords

References

  1. D.Cheng, H.Yan, "Recognition of handwritten digits based on contour information," Pattern Recognition, Vol. 31, No. 3, pp. 235-255, Mar. 1998. https://doi.org/10.1016/S0031-3203(97)00046-0
  2. G.Vamvakas, B.Gatos, S.J.Perantonis, "Handwritten character recognition through two-stage foreground sub-sampling," Pattern Recognition, Vol. 43, No. 8, pp. 2807-2816, Aug. 2010. https://doi.org/10.1016/j.patcog.2010.02.018
  3. N.M.Herbst, C.N.Liu, "Automatic signature verification based on accelerometry," IBMJ. Res. Dev. 21, pp. 245-254, May. 1977. https://doi.org/10.1147/rd.213.0245
  4. G.L.Cash, M.Hatamian, "Optical character recognition by the method of moments," Computer Vision, Graphics, and Image Processing, Vol. 39, pp. 291-310, Sep. 1987. https://doi.org/10.1016/S0734-189X(87)80183-4
  5. K.Fukushima, "Neural NetworkModel for Selective attention in visual pattern recognition and associative recall," Applied Optics, Vol. 26, No. 23, pp. 4985-4992, Dec. 1987. https://doi.org/10.1364/AO.26.004985
  6. M.T.Y.Lai and C.Y.Suen, "Automatic recognition of characte rs by Fourier descriptors and boundary encoding," Pattern Recognition, Vol. 14, pp.383-393, Dec. 1981. https://doi.org/10.1016/0031-3203(81)90083-2
  7. C.L.Giles and T.Maxwell, "Learning, invariance, and generali zation in high-order neural networks," Applied Optics, Vol. 25, pp. 4972-4978, Dec. 1987.
  8. P.Morrison, J.J.Zou, "Triangle refinement in a constrained Delaunay triangulation skeleton," Pattern Recognition, Vol. 40, No. 10, pp 2754-2765, Oct. 2007. https://doi.org/10.1016/j.patcog.2006.12.021
  9. T.F.Smith, M.S.Waterman, "Identification of common molecu lar subsequences," J. Mol. Biol. 147, pp.195-197, Mar. 1981. https://doi.org/10.1016/0022-2836(81)90087-5
  10. C.Xu, J.Liu, X.Tang, "2D Shape Matching by Contour Flex ibility," IEEE Trans. on pattern analysis and machine intelligence, Vol. 31, No. 1, pp. 180-186, Jan. 2009. https://doi.org/10.1109/TPAMI.2008.199