Rough set approach to online signature identification

This paper presents an online signature identification system based on global features. The information is extracted as time functions of various dynamic properties of the signatures. A database of 2160 signatures from 108 subjects was built. Thirty-one features were identified and extracted from each signature. Different feature reduction approaches and classifiers were used to assess their suitability for this application. Rough set approach has resulted in a reduced set of nine features that were found to capture the essential characteristics required for signature identification. Rough set classifier has achieved 100% correct classification rate, which demonstrates its suitability and effectiveness for online signature identification.