ONLINE SIGNATURE VERIFICATION SYSTEM ON MOBILE DEVICES FOR EFFECTIVE AND SECURE BIOMETRIC

Abstract

Handwritten signature is that the most generally accepted biometric to biometric identification. The projected on-line written signature verification system consists principally of 3 phases: Signal preprocessing, feature extraction, and have matching. Steps for confirming on-line written signature during this system begin with extracting dynamic knowledge (x and y positions) of points that forming the signature. Pen-movement angles and speed square measure then derived from pen position knowledge. To scale back variations in pen-position and pen-movement angles spatial property, knowledge is normalized. Options of the signature will be extracted victimization projected feature extraction methodology. Such as each signature a novel feature are extracted and this can be amount victimization quantization step size vector. To verify the check signature Manhattan distance (score) are taken between the check signature and therefore the reference signature. If score is a smaller amount than the predefined threshold then the check signature same is claimed to be real signature and if score is over the predefined threshold then it's said to be cast signature.