Offline Signature Verification for Arabic Language

Biometrics relies on biological features (e.g. finger print, iris or the retina) or behavioral features (voice, signature). Those features can be used for identity verification for an individual. For this it became one of the most trusted and natural ways to identify a person and controlling access to the systems. Signature is a behavioral biometric. Signature is not unique like iris or finger print as it can be forged. Automatic signature verification is divided into two areas depending on the way of data capturing: offline and online signature verification. In offline signature verification, the signature is scanned from a document using a scanner to get the image of the signature. In online signature, a digitizing tablet is used to collect the movements during the signing. In this work we present a system for offline signature verification. In this system the user has to submit a number of signatures which are used to extract two types of features, statistical features and structural features. A vector obtained from each of them is used to train propagation neural net in the verification stage. A test signature is then taken from the user, to compare it with those the net had been trained with. A test experiment was carried out with two sets of data are collected. One set is used as a training set for the propagation neural net in its verification stage. This set with four signatures form each user is used for the training purpose. The second set consisting of one sample of signature for each of the 20 persons is used as a test set for the system. A negative identification test was carried out using a signature of one person to test others’ signatures. The system gave encouraging results