Title :
Signature Recognition Based on Voting Schemes
Author :
Doroz, Rafal ; Porwik, Piotr ; Wrobel, Krzysztof
Author_Institution :
Inst. of Comput. Sci., Univ. of Silesia, Sosnowiec, Poland
Abstract :
In this paper a signature identification system is proposed, where windows technique and voting method have been introduced. The windows technique allows comparing signatures without the algorithms equalizing the lengths of the sequences being compared. The biometric signature recognition system works on the basis of election-voting scheme. The identified signature is compared with signatures stored in database and decision is made by voting system. In the proposed approach dynamic signature features have been recorded, hence this method is preferred for on-line biometric systems. In the method a different signature features and similarity measures can be used. Practical experiments were also conducted with the MCYT signatures´ database and confirmed that results obtained are promising.
Keywords :
digital signatures; feature extraction; government data processing; handwriting recognition; image matching; politics; MCYT signatures database; biometric signature recognition system; dynamic signature features; election-voting scheme; online biometric systems; signature identification system; similarity measures; voting method; voting system; windows technique; Computers; Databases; Hidden Markov models; Iris recognition; Pattern recognition; Sorting; biometrics; signature recognition; similarity measure; voting scheme; windows technique;
Conference_Titel :
Biometrics and Kansei Engineering (ICBAKE), 2013 International Conference on
Conference_Location :
Tokyo
DOI :
10.1109/ICBAKE.2013.11