Abstract

Biometrics authentication has become a recent trend to prevent unauthorized accesses to all kinds of e-data. Signature is strongly accepted in legally and socially as identification and authentication of a person’s identity. But, it is very difficult to verify the signature physically. So, it is needed to design a system that verifies the signature of a human automatically. By putting a single factor authentication together with a 2 factor authentication using signature, our system aim to provide a multifactor authentication system that is more secure and reliable. Using a machine learning model based on neural network, the system is trained and it is able to detect genuine signatures from forgeries.

CHAPTER ONE

GENERAL INTRODUCTION

For centuries, security has always been a priority for mankind. Through letter protection or access to places by particular users, authentication has always been a concrete aspect of security. With the advent of digital nowadays, security remains a major challenge concerning protection of data of different users. From a simple password to the use of biometric parameters, authentication methods evolve over time. In this study, we will study one of the biometric methods of authentication: authentication through signature verification.

1.1 Background and Context of the Study

1.1.1 Background to the Study

A signature is a widely accepted means of authentication in government, legal, and commercial transactions and this from decades. Sumerians were using intricate seals applied to clay cuneiform tablets to authenticate their writings. Documents were authenticated in the Roman Empire by affixing handwritten signatures to the documents. In 1677, England passed an act to prevent frauds and perjuries by requiring documents to be signed by the participating parties. Signature authentication systems aim to confirm the identity of a person based on their signature, that is, they classify signature samples as “genuine” (created by the claimed individual) or “forgery” (created by an impostor). Signature is one of the most popular and commonly accepted biometric hallmarks that have been used since the ancient times for verifying different entities related to human beings: documents, forms, bank checks, individuals, etc. Therefore, signature authentication is a critical task and many efforts have been made to remove the uncertainty involved in the manual authentication procedure, which makes signature authentication an important research line in the field of machine learning and pattern recognition.

1.1.2 Context of Study

In the authentication world, we have what we call factor. A factor is a type of authentication. When you claim to be someone, you need to provide further information to prove that you are who you say you are. Depending on the types of information which are used for authentication purposes we can distinguish many factors of authentication.

One Factor Authentication: Something you know

Information is classified as something you know if you store it in your memory and can retrieve it when needed, for instance, a password, an answer to a security question or a Personal Identification Number (PIN). Also called Single Factor Authentication (SFA), in this type of authentication, the user is asked for that information and depending on its correctness, the user can be authenticated or not.

1.2 Problem Statement

A signature is a mark allowing to identify the author of a document, of a work or the cause of a phenomenon: thus an author signs his writings. A signature can also be affixed at the end of a document by a person to signify his approval of all the information contained in a document of which he is not necessarily the author.

However, some individuals (impostor) may falsify a signature to sign a document in order to pretend to be the author of the document or to signify his/her approval of all the information contained in the document. So it is therefore a question of verifying using machine learning and neural network whether a given signature is genuine (produced by claimed individual) or forgery (produced by and impostor).

1.3 Objectives of the Study

This study has several objectives which can be classified into two categories which are:

1.3.1    General Objective

The general objectives allowing to have ideas centered on the signature and which are:

1. Present the different types of signatures
2. Respectively present the methods of verifying the authenticity of the different types of signatures
3. Differentiate between offline and online signature.

1.3.2 Specific Objectives

The specific objectives are centered on the different techniques and operations that we will use to show that a signature is authentic.

1. Implement neural network technology
2. Use machine learning and neural network to show the authenticity of a signature.
3. Build a multi factor system mobile application