Análise de características locais e globais de assinaturas dinâmicas

Abstract

This study explored digital dynamic signatures containing quantifiable dynamic data and propose a two-step approach with the aim of assessing the potential of biometric data in classifying simulated samples and disguises, using a set of genuine signatures as reference. Thirty individuals contributed voluntarily with a total of 1800 natural, 60 disguises, and 870 simulations samples, including a set of legible, mixed and stylized signatures. Quantitative data analysis tools such as Principal Component Analysis (PCA), boxplots, distance test, and Kolmogorov-Smirnov (KS) test were employed to analyze 62 global characteristics, provided by Wacom Signature Scope software. The PCA analysis was able to correctly group 97,8% of simulated signatures. The study of local characteristics (horizontal and vertical positions, pressure, velocity, acceleration, and jerk) used visual analysis, Dynamic Time Warping (DTW), and the significance of DTW cost differences evaluated through KS test. The classification performance using local characteristics was compared to investigate which local characteristics are more relevant for examination and signature classification. The approach proposed, starting with global features followed by local features, yielded promising results in classifying simulated samples. Disguises could not be satisfactorily differentiated from simulated samples using the proposed approach, consistent with observations in traditional forensic handwriting exam. The error rate for disguises was above 50%, and formal disguises (‘auto simulation’) exhibiting a lower error rate compared to random disguises (‘free-form disguise’). The local characteristics that demonstrated greater robustness in classifying samples were pressure and velocity, warranting further in-depth investigation for possible correlation between them.

References

  1. M. Angel.; J. S. Kelly. Forensic Document Examination in the 21st Century. CRC Press, United States of America. 157-303 (2021).
  2. H. H. Harralson. Developments in Handwriting and Signature Identification in the Digital Age. Routledge, United Kingdom. (2013).
  3. H. Harralson; L. Miller. Huber and Headrick’s Handwriting Identification. CRC press, United States of America. (2017).
  4. J. Linden; R. Marquis; S. Bozza; F. Taroni. Dynamic signatures: A review of dynamic feature variation and forensic methodology, Forensic Sci Int., 291, 216–229 (2018)
  5. J. Linden; R. Marquis; W. Mazzella. Forensic Analysis of Digital Dynamic Signatures: New Methods for Data Treatment and Feature Evaluation. J Forensic Sci., 62, 382–391 (2017).
  6. J. Heckeroth; E. Kupferschmid; T. Dziedzic; N. Kalantzis; B. G. Čakovská; C. Fernandes; P. Schmidt. Features of digitally captured signatures vs. pen and paper signatures: Similar or completely different?. Forensic Sci. Int., 318, 110587 (2021).
  7. M. Caligiuri, L. Mohammed. The neuroscience of handwriting. EPUB format, CRC Press, ISBN 9781466559189 (2012)..
  8. J.B. Okado, M.C. Diana, Análise de assinaturas estáticas produzidas em tela, Rev. Bras. Crimin. 12 (2), 41-46 (2023). Retirado em 14/06/23, de https://doi.org/10.15260/rbc.v12i2.695.
  9. European Network of Forensic Science Institutes (ENFSI). Best Practice Manual for the Forensic Handwriting Examination- Appendix 6 (2022). Retirado em 11/04/2023, de https://enfsi.eu/wp-content/uploads/2023/02/BPM-Handwriting-Ed.-4.pdf.
  10. W. J. Flynn. Conducting a forensic examination of electronically captured signatures. D-ABFDE, 15(1) (2012).
  11. R. Doroz, P. Porwik, T. Para, K. Wrobel. Dynamic signature recognition based on velocity changes of some features. Int. J. Biom., 1(1), 47-62 (2008).
  12. S. N. Srihari, S. H. Cha, H. Arora., S. Lee. Individuality of handwriting. J Forensic Sci., 47(4), 856-872 (2002).
  13. I. W. Evett, C. E. H. Berger, J. S. Buckleton, C. Champod, G. Jackson. Finding the way forward for forensic science in the US—A commentary on the PCAST report. Forensic Sci. Int., 278, 16-23 (2017).
  14. Aitken, C., & Taroni, F. Statistics and the evaluation of evidence for forensic scientists. John Wiley & Sons, United States of America. (2004).
  15. Suporte Wacom. Manual do usuário - Definições de Medidas SignatureScope. (2022).
  16. K. Ahmed; I.M. El-Henawy, M. Z. Rashad; O. Nomir. On-line signature verification based on PCA feature reduction and statistical analysis. International Conference on Computer Engineering & Systems. 3–8 (2010).
  17. L.C. Lee; A.A. Jemain. On overview of PCA application strategy in processing high dimensionality forensic data. Microchem. J., 169, 106608 (2021).
  18. T. Lelore; H. Mouchère; E.M. Moussa. Point to segment distance DTW for online handwriting signals matching, ICPRAM 2023. 1–7 (2023).
  19. K. A. Nicolaides. Using acceleration/deceleration plots in the forensic analysis of electronically captured signatures. D-ABFDE, 15(2). (2012).
  20. M.P. Caligiuri; L.A. Mohammed; B. Found; D. Rogers. Nonadherence to the isochrony principle in forged signatures, Forensic Sci Int. 223, 228–232 (2012).
Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Copyright (c) 2024 Brazilian Journal of Criminalistics

Share

Download

pdf (Português (Brasil))

Author(s)

Most read articles by the same author(s)