ETRI Journal

Electronics and Telecommunications Research Institute (한국전자통신연구원)
권호 표지
DOI QR코드

DOI QR Code

Writer verification using feature selection based on genetic algorithm: A case study on handwritten Bangla dataset

  • Jaya Paul (Department of Computer Science & Engineering, Jadavpur University) ;
  • Kalpita Dutta (Department of Computer Science & Engineering, Jadavpur University) ;
  • Anasua Sarkar (Department of Computer Science & Engineering, Jadavpur University) ;
  • Kaushik Roy (Department of Computer Science, West Bengal State University) ;
  • Nibaran Das (Department of Computer Science & Engineering, Jadavpur University)
  • Received : 2023.05.09
  • Accepted : 2023.11.22
  • Published : 2024.08.20
Download PDF
⟨ Previous Next ⟩

Abstract

Author verification is challenging because of the diversity in writing styles. We propose an enhanced handwriting verification method that combines handcrafted and automatically extracted features. The method uses a genetic algorithm to reduce the dimensionality of the feature set. We consider offline Bangla handwriting content and evaluate the proposed method using handcrafted features with a simple logistic regression, radial basis function network, and sequential minimal optimization as well as automatically extracted features using a convolutional neural network. The handcrafted features outperform the automatically extracted ones, achieving an average verification accuracy of 94.54% for 100 writers. The handcrafted features include Radon transform, histogram of oriented gradients, local phase quantization, and local binary patterns from interwriter and intrawriter content. The genetic algorithm reduces the feature dimensionality and selects salient features using a support vector machine. The top five experimental results are obtained from the optimal feature set selected using a consensus strategy. Comparisons with other methods and features confirm the satisfactory results.

Keywords

References

  1. T. Bahram, A texture-based approach for offline writer identification, J. King Saud Univers. - Comput. Inform. Sci. 34 (2022), no. 8, Part A, 5204-5222.
  2. V. Aubin, M. Mora, and M. Santos Peas, Off-line writer verification based on simple graphemes, Pattern Recogn. 79 (2018), 414-426.
  3. J. A. Lewis, Forensic document examination: fundamentals and current trends, 2014, pp. 1-212.
  4. R. Fernandez-de Sevilla, F. Alonso-Fernandez, J. Fierrez, and J. Ortega-Garcia, Forensic writer identification using allographic features, (23th International Conference on frontiers in Handwriting Recognition, Kolkata, India), 2010, pp. 308-313.
  5. F. J. Zareen and S. Jabin, Authentic mobile-biometric signature verification system, IET Biometrics 5 (2016), no. 1, 13-19, DOI 10.1049/iet-bmt.2015.0017.
  6. V. Aubin, M. Mora, and M. Santos, A new approach for writer verification based on segments of handwritten graphemes, Logic J. IGPL 30 (2022), no. 6, 965-978.
  7. A. Bensefia and H. Tamimi, Validity of handwriting in biometric systems, (Proceedings of the International Conference on Pattern Recognition and Artificial Intelligence, Union, NJ, USA), 2018, pp. 5-10.
  8. C. Halder, S. M. Obaidullah, J. Paul, and K. Roy, Writer verification on Bangla handwritten characters, Vol. 2, Springer, India, 2016.
  9. M. A. Khan, N. Mohammad, G. B. Brahim, A. Bashar, and G. Latif, Writer verification of partially damaged handwritten arabic documents based on individual character shapes, Peer J. Comput. Sci. 8 (2022), e955.
  10. S. He and L. Schomaker, Writer identification using curvature-free features, Pattern Recogn. 63 (2017), no. C, 451-464.
  11. M. S. Obaidullah, C. Halder, C. K. Santosh, N. Das, and K. Roy, Phdindic_11:page-level handwritten document image dataset of 11 official indic scripts for script identification, Multimed. Tools Appl. 77 (2018), 1643-1678.
  12. C. Adak, B. B. Chaudhuri, and M. Blumenstein, An empirical study on writer identification and verification from intra-variable individual handwriting, IEEE Access 7 (2019), 24738-24758.
  13. S. Zhao, C. Zhang, and Y. Wang, Lithium-ion battery capacity and remaining useful life prediction using board learning system and long short-term memory neural network, J. Energy Storage 52 (2022), 104901, DOI 10.1016/j.est.2022.104901.
  14. C. Zhang, S. Zhao, Z. Yang, and Y. Chen, A reliable data-driven state-of-health estimation model for lithium-ion batteries in electric vehicles, Front. Energy Res. 10 (2022), 1013800.
  15. C. Zhang, S. Zhao, Z. Yang, and Y. He, A multi-fault diagnosis method for lithium-ion battery pack using curvilinear manhattan distance evaluation and voltage difference analysis, J. Energy Storage 67 (2023), 107575, DOI 10.1016/j.est.2023.107575.
  16. J. Paul, A. Sarkar, N. Das, and K. Roy, HOG and LBP based writer verification, in Proceedings of international conference on frontiers in computing and systems, D. Bhattacharjee, D. K. Kole, N. Dey, S. Basu, and D. Plewczynski (eds), Springer Singapore, 2021, pp. 3-12, DOI 10.1007/978-981-15-7834-2_1.
  17. J. Miller, R. Patterson, D. Gantz, C. Saunders, M. Walch, and J. Buscaglia, A set of handwriting features for use in automated writer identification, J. Forensic Sci. 62 (2017), 722-734.
  18. S. Bilan, R. Motornyuk, S. Bilan, and O. Galan, User identification using images of the handwritten characters based on cellular automata and radon transform, Biometric Identif. Technol. Based Modern Data Mining Methods 2021 (2021), 91-103.
  19. S. Karanwal, Robust local binary pattern for face recognition in different challenges, Multimed. Tools Appl. 81 (2022), no. 20, 29405-29421.
  20. C.-K. Tran, P. Khamphoui, et al., Face recognition technology using the fusion of local descriptors, Ann. Comput. Sci. Inform. Syst. 34 (2022), 227-231.
  21. W. Siedlecki and J. Sklansky, A note on genetic algorithms for large-scale feature selection, Pattern Recogn. Lett. 10 (1989), no. 5, 335-347.
  22. O. Babatunde, Zernike moments and genetic algorithm: tutorial and application, British J. Math. Comput. Sci. 4 (2014), 2217-2236.
  23. S. Lu, Z. Lu, and Y.-D. Zhang, Pathological brain detection based on AlexNet and transfer learning, J. Comput. Sci. 30 (2019), 41-47.
  24. A. Jain, P. Flynn, and A. Ross, Handbook of biometrics, Springer Science & Business Media, 2008.
  25. H. Hassen and S. Al-Maadeed, Arabic handwriting recognition using sequential minimal optimization, (1st International Workshop on Arabic Script Analysis and Recognition, Nancy, France), 2017, pp. 79-84.
  26. B. Zebardast and I. Maleki, A new radial basis function artificial neural network based recognition for kurdish manuscript, Int. J. Appl. Evol. Comput. (IJAEC) 4 (2013), no. 4, 72-87.
  27. J. Friedman, T. Hastie, and R. Tibshirani, Additive logistic regression: a statistical view of boosting (with discussion and a rejoinder by the authors), Ann. Statist. 28 (2000), no. 2, 337-407.
  28. N. Das, R. Sarkar, S. Basu, M. Kundu, M. Nasipuri, and D. K. Basu, A genetic algorithm based region sampling for selection of local features in handwritten digit recognition application, Appl. Soft Comput. 12 (2012), no. 5, 1592-1606.
  29. C. Halder and K. Roy, Individuality of isolated bangla characters, (International Conference on Devices, Circuits and Communications, Ranchi, India), 2014, pp. 1-6.
  30. R. Hanusiak, L. Soares de Oliveira, E. Justino, and R. Sabourin, Writer verification using texture-based features, Document Anal. Recog. - IJDAR 15 (2011), 1-14.
  31. M. N. Abdi and M. Khemakhem, A model-based approach to offline text-independent arabic writer identification and verification, Pattern Recogn. 48 (2015), no. 5, 1890-1903.