A Comprehensive Study on Offline Signature Verification

Abstract

Handwritten Signatures are special types of behavioural biometric which are used in many applications such as banks, credit cards, passport, check processing, and financial documentation, etc. Verification of these signatures is a challenging task especially in the case of offline where there is no information of signing process. So there is a need for a system that can distinguish between the genuine and the forged signature to avoid the chances of theft or fraud. Many types of researches have been done in this area in the last three decades. Earlier this task was performed by handcrafted features and recently deep learning techniques have been employed for this task, but still, there is a chance of enhancement in the accuracy of the system. In this paper, we present a comprehensive study of the work done in the field of offline signature verification and also the challenges which are still present in this area.

A Hybrid Method of Feature Extraction for Signatures Verification Using CNN and HOG a Multi-Classification Approach

“…Offline signature verification system include two main phases feature extraction and classification, many studies [17,18] indicated that the feature extraction phase effect the signature verification system performance, given its importance and its main role in contributing to facilitating the work of works to distinguish between original signatures and forged signatures, where providing important features contributes to improving and developing the ability of signature verification systems and increasing their degree of accuracy. Despite this, the features extraction stage still faces many issues, such as the quality and number of features extracted from signature images, and this was confirmed by previous studies [19,17,20,21], which indicated the need to improve the features extraction stage to reach a set of features capable of assisting classifiers in verifying signatures.…”

“…In order to make 24 fake signatures for each real signer, each forger made an eight-time effort to copy the signatures of three different signers. As a consequence, there are 1,320 legitimate signatures and 1,320 fake signatures in the dataset (55 24) [20].…”

A Comparative Study among Handwritten Signature Verification Methods Using Machine Learning Techniques

“…Mohammed et al [9] presented a state-of-the-art the methods that were used for capturing signature data as well as methods and techniques that were used in preprocessing, feature extraction, and verification of handwritten signatures. Sharma et al [10] presented a comprehensive study on offline signature verification as well as the challenges in that field. Nehal and Heba [11] presented a comparative study of recent off-line and online identification and verification systems; also discussed the stages of verification and identification systems.…”

Siamese-Network Based Signature Verification using Self Supervised Learning

“…The main purpose of biometrics technology is to provide a means of identifying an individual based on certain merit such as behavioral or physiological attributes. In many cases, this can be achieved by measuring certain aspects such as fingerprint, iris, palm, voice, facial expression, etc [1]. Amongst them, the handwritten signature also can be used to verify certain individuals and is probably still the most commonly used since the ancient times for so many things such as bank checks, forms, insurance, ◼ ISSN (print): 1978-1520, ISSN (online): 2460-7258 IJCCS Vol.…”

“…In the case of handwritten signatures, we can further divide the categories into two major groups: 1) Online Signature Verification and 2) Offline Signature Verification [1], figure 1 for example. Online Signature Verification happens when the writer is writing their signature and involves a special device to record the pen activity such as pressure, speed of writing, angle, coordinate point, etc thus having more complete data available to be used for determining whether the signature is a forgery or genuine than the Offline counterpart.…”