Extraction and Analysis of Document Examiner Features from Vector Skeletons of Grapheme ‘th’

  • Vladimir Pervouchine
  • Graham Leedham
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3872)

Abstract

This paper presents a study of 25 structural features extracted from samples of grapheme ‘th’ that correspond to features commonly used by forensic document examiners. Most of the features are extracted using vector skeletons produced by a specially developed skeletonisation algorithm. The methods of feature extraction are presented along with the results. Analysis of the usefulness of the features was conducted and three categories of features were identified: indispensable, partially relevant and irrelevant for determining the authorship of genuine unconstrained handwriting. The division was performed based on searching the optimal feature sets using the wrapper method. A constructive neural network was used as a classifier and a genetic algorithm was used to search for optimal feature sets. It is shown that structural micro features similar to those used in forensic document analysis do possess discriminative power. The results are also compared to those obtained in our preceding study, and it is shown that use of the vector skeletonisation allows both extraction of more structural features and improvement the feature extraction accuracy from 87% to 94%.

Keywords

Genetic Algorithm Feature Subset Feature Subset Selection Stroke Width Optimal Feature Subset 
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.

References

  1. 1.
    Hilton, O.: Scientific Examination of Questioned Documents. CRC Hall, Florida (1993)Google Scholar
  2. 2.
    Robertson, E.W.: Fundamentals of Document Examination. Nelson-Hall, Illinois (1991)Google Scholar
  3. 3.
    Harrison, W.R.: Suspect Documents, Their Scientific Examinations. Nelson-Hall, Illinois (1981)Google Scholar
  4. 4.
    Lindblom, B.: Document examination. In: Chayko, G.M., Gulliver, E.D. (eds.) Forensic Evidence in Canada, 2nd edn., pp. 505–525. Canada Law Book, Aurora (1999)Google Scholar
  5. 5.
    Daubert et al. v. Merrell Dow Pharmaceuticals. Court case 509 U. S. 579 (1993)Google Scholar
  6. 6.
    Kam, M., Fielding, G., Conn, R.: Writer identification by professional document examiners. Journal of Forensic Sciences 42, 778–786 (1997)Google Scholar
  7. 7.
    Found, B., Sita, J., Rogers, D.: The development of a program for characterizing forensic handwriting examiners’ expertise: Signature examination pilot study. Journal of Forensic Document Examination 12, 69–80 (1999)Google Scholar
  8. 8.
    Kam, M., Gummadidala, K., Fielding, G., Conn, R.: Signature authentication by forensic document examiners. Journal of Forensic Science 46, 884–888 (2001)Google Scholar
  9. 9.
    Sita, J., Found, B., Rogers, D.: Forensic handwriting examiners’ expertise for signature comparison. Journal of Forensic Sciences 47, 1117–1124 (2001)Google Scholar
  10. 10.
    Srihari, S.N., Cha, S.H., Arora, H., Lee, S.: Individuality of handwriting. Journal of Forensic Sciences 47, 1–17 (2002)Google Scholar
  11. 11.
    Found, B., Roger, D., Schmittat, R.: A computer program designed to compare the spatial elements of handwriting. Forensic Science International 68, 195–203 (1994)CrossRefGoogle Scholar
  12. 12.
    Schomaker, L., Bulacu, M., van Erp, M.: Sparse-parametric writer identification using heterogeneous feature groups. In: Proc. 3rd Int’l Conf. Document Analysis and Recognition (ICDAR 1995), Montreal, Canada, pp. 545–548 (1995)Google Scholar
  13. 13.
    Srihari, S.N., Cha, S.H., Lee, S.: Establishing handwriting individuality using pattern recognition techniques. In: Proc. 6th Int’l Conf. Document Analysis and Recognition (ICDAR 2001), Seattle, USA, pp. 1195–1204 (2001)Google Scholar
  14. 14.
    Tomai, C.I., Zhang, B., Srihari, S.N.: Discriminatory power of handwritten words for writer recognition. In: Proc. 17th Int’l Conf. Pattern Recognition, Cambridge, UK, pp. 638–641 (2004)Google Scholar
  15. 15.
    Zhang, B., Srihari, S.N., Lee, S.: Individuality of handwritten characters. In: Proc. 7th Int’l Conf. Document Analysis and Recognition (ICDAR 2003), Edinburgh, UK, pp. 1086–1090 (2003)Google Scholar
  16. 16.
    Srihari, S.N., Tomai, C.I., Zhang, B., Lee, S.: Individuality of numerals. In: Proc. 7th Int’l Conf. Document Analysis and Recognition (ICDAR 2003), Edinburgh, UK, pp. 1096–1100 (2003)Google Scholar
  17. 17.
    Huber, R.A., Headrick, A.M.: Handwriting Identification: Facts and Fundamentals. CRC Press, LCC (1999)CrossRefGoogle Scholar
  18. 18.
    Leedham, C.G., Pervouchine, V., Tan, W.K., Jacob, A.: Automatic quantitative letter-level extraction of features used by document examiners. In: Teulings, H.L., Van Gemmert, A.W.A. (eds.) Proc. 11th Conf. Int’l. Graphonomics Society (IGS 2003), Scottsdale, AZ, USA, pp. 291–294 (2003)Google Scholar
  19. 19.
    Leedham, C.G., Pervouchine, V., Tan, W.K.: Quantitative letter-level extraction and analysis of features used by document examiners. Journal of Forensic Document Examination (2004) (in press)Google Scholar
  20. 20.
    Leedham, C.G., Pervouchine, V.: Validating the use of handwriting as a biometric and its forensic analysis. In: Pal, U., Parui, S.K., Chaudhuri, B.B. (eds.) Document Analysis: Proc. Int’l. Workshop on Document Analysis (IWDA 2005), pp. 175–192. Allied Publishers Ltd, Chennai (2005) (Invited lecture)Google Scholar
  21. 21.
    Pervouchine, V., Leedham, C.G., Melikhov, K.: Handwritten character skeletonisation for forensic document analysis. In: Proc. 20th Annual ACM Symposium on Applied Computing, Santa Fe, NM, USA, pp. 754–758 (2005)Google Scholar
  22. 22.
    John, G.H., Kohavi, R., Pfleger, K.: Irrelevant features and the subset selection problem. In: Proc. 11th Int’l Conf. Machine Learning (ML 1994), Rutgers University, New Brunswick, NJ, pp. 121–129 (1994)Google Scholar
  23. 23.
    Weiss, S.M., Kulikowski, C.A.: Computer Systems that Learn: Classification and Prediction Methods from Statistics, Neural Nets. In: Machine Learning, and Expert Systems. Morgan Kaufmann, San Francisco (1991)Google Scholar
  24. 24.
    Yang, J., Parekh, R., Honavar, V.: Distal: An inter-pattern distance-based constructive learning algorithm. Technical Report ISU-CS-TR 97-06, Department of Computer Science, Iowa State University (1997); Also appeared in Proc. Int’l. Conf. Neural Networks. IEEE, Piscataway (1998)Google Scholar
  25. 25.
    Vafaie, H., De Jong, K.: Genetic algorithms as a tool for feature selection in machine learning. In: Proc. 4th Int’l Conf. Tools with Artificial Intelligence (TAI 1992), pp. 200–203. IEEE Computer Society Press, Arlington (1992)CrossRefGoogle Scholar
  26. 26.
    Brill, F.Z., Brown, D.E., Martin, W.N.: Fast genetic selection of features for neural network classifiers. IEEE Trans. Neural Networks 3, 324–328 (1992)CrossRefGoogle Scholar
  27. 27.
    Bala, J., Huang, J., Vafaie, H., De Jong, K., Wechsler, H.: Hybrid learning using genetic algorithms and decision trees for pattern classification. In: Proc. Int’l Joint Conf. Artificial Intelligence (IJCAI 1995), Montreal, Canada (1995)Google Scholar
  28. 28.
    Chen, S., Smith, S., Guerra-Salcedo, C., Whitley, D.: Fast and accurate feature selection using hybrid genetic strategies. In: Proc. Congress on Evolutionary Computation (CEC 1999), Washington DC, USA (1999)Google Scholar
  29. 29.
    Yang, J., Honavar, V.: Feature subset selection using a genetic algorithm. IEEE Intell. Syst. 13, 44–49 (1998)CrossRefGoogle Scholar
  30. 30.
    Goldberg, D.E.: Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley, Reading (1989)MATHGoogle Scholar
  31. 31.
    Syswerda, G.: Uniform crossover in genetic algorithms. In: Proc. 3rd Int’l Conf. Genetic Algorithms, George Mason University, USA, pp. 2–9. Morgan Kaufmann, San Francisco (1989)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Vladimir Pervouchine
    • 1
  • Graham Leedham
    • 1
  1. 1.Forensics and Security LabNanyang Technological University, School of Computer EngineeringSingapore

Personalised recommendations