Signature verification and recognition by using support vector machine

Abstract

Signature is an important biometric of a person that is attribute of a human being which can be used to authenticate human identity. There are lots of biometric techniques that had been represented in the past. A biometric‟s signature verification type had play a very important role in financial, legal transaction and commercial, truly method of authenticating becomes more and more crucial. There are various approaches to signature recognition with a lot of scope of researches. In this project, off- line signature recognition and verification using support vector machine (SVM) is proposed. First, feature vectors are extracted for a reference set of signatures using radon transform. Then, the SVM is trained using a set of genuine and a set of forgery signature. Once trained, the SVM is expected to recognize using an input signature.