Prototype for signature verification system using euclidean distance

Abstract

On a daily basis we use signature to authenticate and verify a person's identification. This emphasizes the need for an automatic verification system. The biggest problem in the offline signature verification is the acute lack in representation of signature on the basis of the shape and other features. The main problem is the unpredictability of the shapes of the region of interest i.e. the signature, which is related to the identity of a person. To authenticate a transaction a person's signature is taken as a validation proof. We are proposing a prototype, an offline signature verification system that extracts seven features of a given signature and compares them with the genuine signature contained in the database to see if a particular input signature is of that person or not.