Method For Off-line Signature Verification Research Articles

Identification of Proper and Improper Signatures Using Graph Theory Techniques

Objectives: To propose an automatic signature identification for off-line signature utilising graph theory approaches. Methods: Scanned signatures (Kaggle, https://www.kaggle.com/divyanshrai/handwritten-signatures/data) are collected for off-line signature data. The method follows pre-processing, vertex point extraction by midpoint traverse method, features extraction using edge, average edge and average edge D-distance and Support Vector Machine (SVM) to classify and predict the true label for the genuine and forged signatures. False Acceptance Ratio (FAR) and False Rejection Ratio (FRR) give the accuracy of the proposed methods. This off-line signature verification method is compared with the deep learning techniques existing in the literature. Findings: Support Vector Machine (SVM) used for classification and results on standard signature datasets like ICDAR (International Conference on Document Analysis and Recognition). The results demonstrate how the proposed strategy outperforms the state-of-the-art already available. Novelty: The proposed approach use the edge distance, average edge distance, and average edge D-distance inbuilt graph structures to extract the feature points. Keywords: Signature images; grid approach; bipartite graph; complete bipartite graph; mid point traverse method

Improvement of Offline Signature Verification Method Using Discrete Wavelet Transform

Improvement of Offline Signature Verification Method Using Discrete Wavelet Transform

Uyghur Off-line Signature Verification Based on Modified Corner Line Features

A modified corner line features based off-line signature verification method proposed for Uyghur handwritten signature in this paper. The signature images were preprocessed according to the nature of Uyghur signature. Then 3 types of corner line features and modified corner curve features were extracted separately. Experiments were performed using Euclidean distance classifier and non-linear SVM classifier for Uyghur signature samples from 150 genuine signatures, 72 random and skilled forgeries are selected from our Uyghur handwritten signature database. Experiments indicate that the MCLF-48 with training 75 samples has obtained 2.16% of FRR and 2.27% of FAR with none-linear SVM classifier. It was concluded that modified corner line features based verification method can capture the nature of Uyghur signature and its writing style more efficiently.

基于方向特征的维吾尔文离线签名鉴别

本文以弥补我国少数民族离线签名鉴别体系中这一漏洞的前提下，结合维吾尔文手写签名的书写风格及特点，提出了一种基于16维方向的维吾尔文离线签名鉴别方法。实验中，对10个维吾尔族人的600个手写签名样本进行签名预处理和特征提取之后，分别用三种距离分类器来进行了签名鉴别，得到的最高签名总正确率为92.58%。实验结果表明，本文提出的方法对于离线的维吾尔文手写签名鉴别来说是一种较有效的鉴别方法。 In this paper, in order to compensate the blank of ethnic minorities off-line signature verification system in our country, a Uyghur off-line signature verification method based on 16 dimensional directional features was proposed by combining with the writing style and nature of Uyghur handwritten signature. In the experiment, three kinds of distance classifier were used separately to signature verification after preprocessing and feature extracting 600 handwritten signature samples from the 10 Uyghur people, and 92.58% of the highest total correct verification rate was obtained. Experimental results show that, the method proposed in this paper is a more effective verification method for off-line handwritten Uyghur signature.

New off-line Handwritten Signature Verification method based on Artificial Immune Recognition System

Natural Immune System offers many interesting features that inspired the design of Artificial Immune Systems (AIS) used to solve various problems of engineering and artificial intelligence. AIS are particularly successful in fault detection and diagnosis applications where anomalies such as errors and failures are assimilated to viruses that should be detected. Thereby, AIS seem suitable to automatically detect forgeries in signature verification systems. This paper proposes a novel method for off-line signature verification that is based on the Artificial Immune Recognition System (AIRS). For feature generation, two different descriptors are proposed to generate signature traits. The first is the Gradient Local Binary Patterns that estimates gradient features based on the LBP neighborhood. The second descriptor is the Longest Run Feature, which describes the signature topology by considering longest suites of text pixels. Performance evaluation is carried out on CEDAR and GPDS-100 datasets. The results obtained showed that the proposed system has promising performance and often comfortably outperforms the state of the art.

Off-line signature verification using structural feature correspondence

This paper presents an off-line signature verification method using a model-based approach. In this method, statistical models are constructed for both pixel distribution and structural layout description. In addition to simple geometric handwriting features, it is proposed to use the directional frontier feature as a structural descriptor of the signature. The statistical verification algorithm based on the geometric handwriting feature is used to accept signatures which closely resemble the reference samples, and to reject random and less skilled forgeries. For the questionable signatures for which the pixel feature judgement is inconclusive, the structural feature verification algorithm is invoked. This algorithm compares the detailed structural correlation between the input and reference signatures in an attempt to detect skilled forgeries. The effectiveness of the approach is evaluated on an experimental signature database.

Off-line signature verification based on geometric feature extraction and neural network classification

In this paper a method for off-line signature verification based on geometric feature extraction and neural network classification is proposed. The role of signature shape description and shape similarity measure is discussed in the context of signature recognition and verification. Geometric features of input signature image are simultaneously examined under several scales by a neural network classifier. An overall match rating is generated by combining the outputs at each scale. Artificially generated genuine and forgery samples from enrollment reference signatures are used to train the network, which allows definite training control and at the same time significantly reduces the number of enrollment samples required to achieve a good performance. Experiments show that 90% correct classification rate can be achieved on a database of over 3000 signature images.