This requires the creation and the diversification of new online and offline signature verification methods. The signature verification methods contain both online (or dynamic) and offline (or static) signature verification methods. In this paper, an offline digital signature verification technique is proposed, that depends on extracting several features from the signatures to be used during simulation. | International Journal of Computer Networks and Communications Security C VOL. 1, NO. 4, SEPTEMBER 2013, 143–151 Available online at: ISSN 2308-9830 N C S Assessment of Offline Digital Signature Recognition Classification Techniques DINA DARWISH1 1 Assisstant professor, International Academy for engineeering and Media Science, 6th October, Egypt E-mail: ABSTRACT The digital signature verification has become an interesting domain, which is widely needed. The usage of online and offline digital signatures has been spreaded worldwide due to the increase of use of bank transactions and user authentication and other similar activities. This requires the creation and the diversification of new online and offline signature verification methods. The signature verification methods contain both online (or dynamic) and offline (or static) signature verification methods. In this paper, an offline digital signature verification technique is proposed, that depends on extracting several features from the signatures to be used during simulation. Some signatures were used for training and others were used for testing only. Different methods such as, vectors manipulation, ensemble classification using boosted trees, and bagged trees, were used in this paper during simulation to obtain results. Keywords: Signature Verification, Offline Digital Signature, Vectors Manipulation, Ensemble Classification, Bagged Trees. 1 INTRODUCTION The growth in today's online and offline transactions that includes banking transactions has posed the question of how to make secure online and offline signature verification techniques, to eliminate the possibility of personal information theft. There are a different number of personnel characteristics that can be used to identify each person, such as, voice, lip movements, hand geometry, face, iris, retina, fingerprint, and others. These characteristics are called biometrics, and these biometrics can be .
