Neural network approach to verifying manual signature

Manual signature is widely used in daily life. There are several different approaches to its recognition. This article deals with the method of static (offline) verification of manual signature. We compare various approaches to verification with regard to the errors of the first and second kinds. The article reviews publications in this topical area and describes general verification algorithm with the use of artificial neutral networks. We consider the suggested verification algorithm using multi-layer neural network, describe its stages, present the results of its implementation in the Python language and make a conclusion concerning the prospects for further development of this field.

References

[1] Dash T., Nayak T., Chattopadhyay S. Handwritten signature verification (offline) using neural network approaches: a comparative study. International Journal of Computer Applications, 2012, vol. 57, no. 7, pp. 33–41.

[2] McCabe A., Trevathan J., Read W. Neural network-based handwritten signature verification. Journal of computers, 2008, vol. 3, no. 8, pp. 9–22.

[3] Coetzer J., Herbst B.M., du Preez J.A. Offline signature verification using the discrete radon transform and a hidden Markov model. EURASIP Journal on Advances in Signal Processing, 2004, vol. 4, pp. 559–571.

[4] Deshmukh A., Desai S., Chaure T., Chothe A., Wankhade S. Automatic signature verification with chain code using weighted distance and Euclidean distance — a review. International Journal of Research in Engineering and Technology, 2016, vol. 5, no. 3, pp. 228–230.

[5] Moolla Y., Viriri S., Nelwamondo F., Tapamo J.S. Offline signature verification using locally optimized distance-based classification. South African Computer Journal, 2013, vol. 50, pp. 15–28.

[6] Daqrouq K., Sweidan H., Balamesh A., Ajour M. Off-line handwritten signature recognition by wavelet entropy and neural network. Entropy, 2017, vol. 19, no. 6, art. 252.

[7] Khan S., Dhole A. An offline signature recognition and verification system based on neural network. International Journal of Research in Engineering and Technology, 2014, vol. 3, no. 11, pp. 443–448.

[8] Lozhnikov P.S., Sulavko A.E., Eremenko A.V., Volkov D.A. Experimental evaluation of reliability of signature verification by quadratic form networks, fuzzy extractors and perceptrons. Informatsionno-upravlyayushchie sistemy [Information and Control Systems], 2016, no. 5, pp. 73–85.

[9] Azzopardi G. How effective are radial basis function neural networks for offline handwritten signature verification? Thesis, B. Sc. University of London, 2006. 123 p.

[10] Petrenko S. Eto nuzhno znat’: Klyuchevye rekomendatsii po glubokomu obucheniyu (Chast’ 2) [What you need to know: key recommendations for deep learning (Part 2)]. Available at: http://datareview.info/article/eto-nuzhno-znat-klyuchevyie-rekomendatsii-po-glubokomu-obucheniyu-chast-2/ (accessed 03.11.2017).