Biometric Based Online Signature Identification to Determine the Fraudulent Behavior

Abstract

Signature verification is the promising technique for one’s own personal identification in day to day life activities like banking transactions, credit card payments and all kind of legal documents. Nowadays wifi credit cards offered by banks are in major threat as it doesn’t need any sort of verification. The customer can be verified either through online or offline depending on the input. In this paper, i built self database after collecting the details of 20 persons like their pen speed, pen angle, pen acceleration, x,y coordinates of the pen from the way of writing. Fuzzy logic is used to compute the accuracy of the recognition rate and the implementation is accomplished in Matlab. It is found that the results are remarkable and the accuracy rate is noticeable.