Online handwritten signature verification based on the most stable feature and partition
- 112 Downloads
Existing methods for online signature verification are generally writer independent, as a common set of features is used for all writers during verification. In this paper, we propose a new method of online handwritten signature verification. Our approach is based on the writer dependent feature as well as writer dependent partition. The two decisions namely optimal feature suitable for a writer and a partition to be used for authenticating the writer, they are taken based on the error rate at the training phase. It is difficult for the forger to imitate the shape and dynamic characteristics of the signer at the same time. According to this feature, we propose to decompose signature trajectories depending upon pressure, velocity direction angle, and velocity information and perform verification on the most stable partition. Experimental results demonstrate superiority of our approach in online signature verification in comparison with other schemes.
KeywordsSignature verification The most stable feature Partition
We would like to thanks the anonymous reviewers for their careful reading and useful comments. This work was supported by the National Natural Science Foundation of China (U1405255, 61671360, 61672415, 61672409), and the Fundamental Research Funds for the Central Universities (JB161505).
- 3.Hafemann, L.G., Sabourin, R., Oliveira, L.S.: Analyzing features learned for offline signature verification using deep CNNs. In: International Conference on Pattern Recognition (2016)Google Scholar
- 8.Tahir, M., Akram, M.U., Idris, M.A.: Online signature verification using segmented local features. In: International Conference on Computing, pp. 100–105 (2016)Google Scholar
- 9.Richiardi, J., Drygajlo, A.: Gaussian mixture models for on-line signature verification. In: ACM Workshop on Biometrics Methods and Applications (WBMA), pp. 115–122 (2003)Google Scholar
- 13.Porwik, P., Doroz, R.: Self-adaptive biometric classifier working on the reduced dataset. Comput. Sci. 8480, 377–388 (2014)Google Scholar
- 16.Zalasiński, M., Cpałka, K., Rakus-Andersson, E.: An idea of the dynamic signature verification based on a hybrid approach. Artif. Intell. Soft Comput. Appl. 9, 232–246 (2016)Google Scholar
- 17.Aqili, N., Maazouzi, A., Raji, M.: On-line signature verification using point pattern matching algorithm. In: International Conference on Electrical and Information Technologies, pp. 410–413 (2016)Google Scholar
- 19.Paudel, N., Querini, M., Italiano, G.F.: Handwritten signature verification for mobile phones. In: International Conference on Information Systems Security and Privacy, pp. 46–52 (2016)Google Scholar
- 20.Plamondon, R., Parizeau, M.: Signature verification from position, velocity and acceleration signals: a comparative study. In: IEEE 9th International Conference on Pattern Recognition, vol. 2, pp. 710–717 (1988)Google Scholar
- 21.Hastie, T., Kishon, E., Clark, M., Fan, J.: A model for signature verification. IEEE Int. Conf. Syst. Man Cybern. 1, 600–604 (1991)Google Scholar
- 22.Plamondon, R., Yergeau, P., Brault, J.J.: A multi-level signature verification system. In: From Pixels to Features III—Frontiers in Handwriting Recognition, pp. 363–370. North-Holland, New York (1992)Google Scholar
- 23.Yeung, D.-Y., Chang, H., Xiong, Y., George, S., Kashi, R., Matsumoto, T., Rigoll, G.: SVC2004: first international signature verification competition. In: ICBA 2004, LNCS, vol. 3072, pp. 16–22. Springer, Berlin (2004)Google Scholar