Confidence Measures in Recognizing Handwritten Mathematical Symbols

  • Oleg Golubitsky
  • Stephen M. Watt
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5625)

Abstract

Recent work on computer recognition of handwritten mathematical symbols has reached the state where geometric analysis of isolated characters can correctly identify individual characters about 96% of the time. This paper presents confidence measures for two classification methods applied to the recognition of handwritten mathematical symbols. We show how the distance to the nearest convex hull of nearest neighbors relates to the classification accuracy. For multi-classifiers based on support vector machine ensembles, we show how the outcomes of the binary classifiers can be combined into an overall confidence value.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Char, B., Watt, S.M.: Representing and Characterizing Handwritten Mathematical Symbols through Succinct Functional Approximation. In: Proc. Intl. Conf. on Docum. Anal. and Rec. (ICDAR), pp. 1198–1202 (2007)Google Scholar
  2. 2.
    Golubitsky, O., Watt, S.M.: Online Stroke Modeling for Handwriting Recognition. In: Proc. 18th Intl. Conf. on Comp. Sci. and Soft. Eng. (CASCON), pp. 72–80 (2008)Google Scholar
  3. 3.
    Golubitsky, O., Watt, S.M.: Online Computation of Similarity between Handwritten Characters. In: Proc. Document Recognition and Retrieval (DRR XVI), pp. C1–C10 (2009)Google Scholar
  4. 4.
    Golubitsky, O., Watt, S.M.: Online Recognition of Multi-Stroke Symbols with Orthogonal Series. In: ICDAR (accepted, 2009)Google Scholar
  5. 5.
    Golubitsky, O., Watt, S.M.: Improved Character Recognition through Subclassing and Runoff Elections. Ontario Research Center for Computer Algebra Tecnical Report TR-09-01Google Scholar
  6. 6.
    Golubitsky, O., Watt, S.M.: Tie Breaking for Curve Multiclassifiers. Ontario Research Center for Computer Algebra Tecnical Report TR-09-02Google Scholar
  7. 7.
    Watt, S.M.: Mathematical Document Classification via Symbol Frequency Analysis. In: Proc. Towards Digital Mathematics Library (DML 2008), pp. 29–40 (2008)Google Scholar
  8. 8.
    Li, M., Sethi, I.: Confidence-Based Classifier Design. Pattern Recognition 39(7), 1230–1240 (2006)CrossRefMATHGoogle Scholar
  9. 9.
    Vincent, P., Bengio, Y.: K-local Hyperplane and Convex Distance Nearest Neighbor Algorithms. In: Adv. in Neural Inform. Proc. Systems, pp. 985–992. MIT Press, Cambridge (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Oleg Golubitsky
    • 1
  • Stephen M. Watt
    • 1
  1. 1.University of Western OntarioLondon, OntarioCanada

Personalised recommendations