International Journal of Electronics, Communication &
Instrumentation Engineering Research and
Development (IJECIERD)
ISSN 2249-684X
Vol. 3, Issue 4, Oct 2013, 1-12
© TJPRC Pvt. Ltd.
OFFLINE SIGNATURE VERIFICATION USING STATISTICAL FEATURE
DEVSHRI SATYARTHI & R. K. GUPTA
Department of CSE & IT, M.I.T.S., Gwalior, Madhya Pradesh, India
ABSTRACT
A method for offline (static) handwritten signature identification and verification proposed based on a
combination of different statistical measures. In the proposed method focus is on the combination of statistical measures in
order to improve the overall efficiency as individual measure lack in providing unique feature for different signature and
provide generalized feature for minute change in signature of the same person. The proposed method is tested our own
signature database contains 400 offline signature of individuals including, 1 test signature and the result are compare with
other state of art of method sand prove that proposed methods is better in terms of efficiency.
KEYWORDS: Signature Verification, Statistical Feature, Threshold, Euclidean Distance
INTRODUCTION
Signature is composed of special characters and brandish and therefore most of the time they can be unreadable.
Signature are the primary mechanism both authentication and authorization in legal document and transaction, so need for
research scope of person authentication has increased in upcoming year. Offline signature based on only pixel image can
be evaluated. Offline signature arrest only one time writing process, in which all information available in static images.
One most important is offline has the advantage of using it in the same way as the existing manual recognition method.
Recognition and Verification are most important method to finding the signature is genuine or forgery.
Recognition means to find the identification of the signature owner, and Verification means taking a decision about
whether the signature is genuine or forgery.
In this paper focus on the offline signature verification to based on statistical analysis. In paper take three type
signatures there are: (i) random, (ii) simple, (iii) skilled. Random forgeries are written by that person who doesn‟t know the
shape of original signature. Simple forgeries are written by a person who knows the shape of original signature without
much practice. Skilled forgeries are written by a person who knows the shape of original signature without much practice,
and forgeries are four types there are: (a) genuine signature, (b) random forger, (c) simulated simple forgery, and (d)
simulated skilled forgery. Genuine signature is original signature, random signature is something learn but not original
signature, simulated simple forgery is copy not know original shape of signature without any practice, and simulated
skilled forgery is copy not know original shape of signature with need practice.
QUALITY PERFORMANCE MEASURES
In evaluating the performance of a signature verification system, there are two important factors: the false
rejection rate (FRR) of genuine signatures and the false acceptance rate (FAR) of forgery signatures and these two are
inversely related.
The false rejection rate (FRR), the false acceptance rate (FAR), the equal error rate (EER) and the average rate
(AER) are used as quality performance measures.