A.I. Korolev - Design Engineer, LLC «VED-Servis». E-mail: korolev_ai@mail.ru V.F. Zhirkov - Ph. D. (Eng.), Professor, Department of Computer Engineering, Vladimir State University named after A.&N. Stoletovs. E-mail: jirkovvf@mail.ru

There is dependence of sign recognition quality on Radon transform-based algorithm considered in this article. The modeling system at Matlab environment is shortly described. The free (for non-commercial use) set «SVC2004» of sign images was selected for investigation. There are main points of Radon transform sign recognition algorithm shown. The discretization, normalization and post-processing features for source image and transform results are described. The Equivalent Error Rate was selecting as criterion of quality. The sign recognition quality estimation was differently processed for identification and verification tasks. The interim and final results was shown at table and diagram forms. The sub-optimal parameters for identification and verification was found and described. This sub-optimal parameters using is opportunity for reducing algorithm realization performance requirement and for improving rec-ognition quality.

 

1. Nguyen Vu, Blumenstein M., Leedham G. Global Features for the Off-Line Signature Verification Problem // 10th International Conference on Document Analysis and Recognition. 2009. URL: http://www.cvc.uab.es/icdar2009/papers/3725b300.pdf (21.05.2015).
2. Srikanta Pal, Michael Blumenstein, Umapada Pal. Off-Line Signature Verification Systems: A Survey. URL: http://www98.griffith.edu.au/dspace/bitstream/handle/10072/44743/77275_1.pdf-sequence=1 (data obrashhenija 03.06.2015).
3. «Matlab. Embedded systems». URL: http://www.mathworks.com/embedded-systems. (data obrashhenija 03.06.2015).
4. «SVC 2004: First International Signature Verification Competition». URL: http://www.cse.ust.hk/svc2004 (11.06.2015).
5. Dit-Yan Yeung, Hong Chang, Yimin Xiong, Susan George, Ramanujan Kashi, Takashi Matsumoto, Gerhard Rigoll. SVC2004: First International Signature Verification Competition. URL: http://www.jdl.ac.cn/user/hchang/doc/icba04.pdf (22.05.2015).
6. Guest R.M. «PENTOOLS - A MATLAB Toolkit for On-line Pen-based Data Experimentation».  URL: http://www.sciweavers.org/read/pentools-a-matlab-toolkit-for-on-line-pen-based-data-experimentation-314206 (02.03.2015).
7. Toft P. The Radon Transform: Theory and Implementation.Ph. D. Thesis, Department of Matematical Modelling Section for Digital Signal Processing, Technical University of Danemark. 1996.  URL: http://orbit.dtu.dk/fedora/objects/orbit:85899/datastreams/file_5529668/content (data obrashhenija 14.11.2013).
8. Koroljov A.I., ZHirkov V.F. Ocenka ehffektivnosti algoritma raspoznavanija rukopisnojj podpisi na osnove preobrazovanija Radona // Trudy XIMezhdunar. nauchnojj konf. «Fizika i radioehlektronika v medicine i ehkologii» s ehlementami nauchnojj molodezhnojj shkoly. FREHMEH-2014. 1-3 ijulja 2014. Sekcija 5. Intellektualnye biometricheskie sistemy i tekhnologii (Nauchnaja molodezhnaja shkola). Kniga 2. Vladimir: 2014. ISBN 978-5-905527-08-1. S. 39−42.
9. Koroljov A.I., ZHirkov V.F. Issledovanie vlijanija parametrov preobrazovanija Radona v zadache raspoznavanija rukopisnojj podpisi // Trudy XI Mezhdunar. simpoziuma «Intellektualnye sistemy, INTELS-2014. Rossija. Moskva. RUDN. 30 ijunja - 4 ijulja 2014 / RAN; RAEN; MGTU im.N.EH. Baumana; Raketno-kosmicheskaja korporacija «EHnergija» im.S.P. Koroljova; Vychislitelnyjj centr im.A.A. Dorodnicyna RAN / Pod red. K.A. Pupkova. M.: RUDN. 2014. 552 s. S. 28−32.
10. Blankers V.L., C.E. van den Heuvel, Franke K.Y.Vuurpijl L.G. The ICDAR 2009 Signature Verification Competition. 2009.  URL: http://www.forensischinstituut.nl/Images/idcar-2009-signature-verification-competition_tcm119-447114.pdf (data obrashhenija 27.05.2015).