Enhancements on Off-line Signature Verification and Recognition
Table of Contents
Chapter 1: Introduction3
Background of the study3
Problem Statement5
Purpose of the study7
Research Objectives7
Literature search7
Proposed Algorithms7
Scope8
Significance of the Study8
Chapter 2: Literature Review9
Chapter 3: Methodology15
Introduction15
Steps Used in Offline Handwritten Signature Verification15
Signature Enrolment16
Signature Verification16
Measurement of the Signature Verifier Accuracy16
Comparison with Human Expert16
Chapter 1: Introduction
Background of the study
Handwritten signatures are widely accepted as a means of document authentication, authorization and personal verification. For legality most documents like bank cheques, travel passports and academic certificates need to have authorized handwritten signatures. In modern society where fraud is rampant, there is the need for an automatic HSV(Handwritten signature verification) system to complement visual verification. Automated signature verification is as important as other automatic identification systems, though they differ from other systems that rely on possession of keys e.t.c or knowledge of specific personal information like passwords. They rely on well learned gestures and still they are most socially and legally accepted form of personal identification [3, 4]. Biometrics can be classified into two types; physiological and behavioural. Physiological biometrics measure some physical features of the subject like fingerprints, iris, hand and finger geometry which are stable over time. Behavioural biometrics measures user actions like speaking, writing and walking which are affected by health, age and physiological factors [5, 6]. A signature is a behavioural biometric characterised by behavioural trait that a writer learns and acquires over a period of time and becomes his unique identity [7, 5]. HSV systems are suited for forgery detection as they are cheap and nonintrusive and provide a direct link between the writer's identity and the transaction [1]. The objective of signature verification systems is to differentiate between original and forged signature, which is related to intra-personal and inter-personal variability [8, 9]. Intra-personal variation is variation among the signatures of the same person and inter-personal is the variation between the originals and the forgeries [9, 7]. We make a distinction between signature recognition and signature verification. Verification decides whether a claim that a particular signature belong to a specific class (writer) is true or false whereas recognition decides to which of a certain number of classes(writers) a particular signature belongs [1, 10, 5]. Automatic HSV systems are classified into two: offline HSV and online HSV [11, 12]. The online signature is captured using a special pen called a stylus and digitizing tablet and analysis is based on dynamic characteristics like pressure, velocity, acceleration and capture time of each point on the signature trajectory. In offline systems the input is a static image that is scanned and used for analysis. Both offline and online systems are used to detect various types of forgeries. Signature forgeries are classified as follows [8, 13, 1, 4, and 12]:
Random/simple or zero effort. The forger doesn't have the shape of the writer signature but comes up with a scribble of his own. He may derive this from the writer's name. This forgery accounts for majority of forgery cases though it's easy to detect with naked ...