Showing papers on "Signature recognition published in 2019"

Reducing the template ageing effect in on-line signature biometrics

Abstract: On-line signature recognition is an area of growing interest in recent years due to the massive deployment of high-quality digitising tablets, smartphones, and tablets in many commercial sectors such as banking. In addition, handwritten signature is one of the most socially accepted biometric traits as it has been used in financial and legal agreements for over a century. In this current environment for signature biometrics, the number of stored samples or templates per user can grow very fast, making it possible to train more robust statistical user models, improving the performance of the biometric systems and in particular reducing the template ageing effect. This study carries out an exhaustive experimental analysis of template update strategies for three well-known on-line signature verification approaches, extracts various practical findings related to the template ageing effect in signature biometrics, and configures time-adaptive improved versions of the considered baseline approaches overcoming to some extent the template ageing. The proposed improved approach achieves system performances of 2.1 and 0.2% equal error rate for skilled and random forgery cases, respectively. These results show the efficacy of the proposed methodology.

Quantifying dynamic time warping distance using probabilistic model in verification of dynamic signatures

Abstract: One of the multimodal biometric scenarios is realized by considering several features coming from a single biometric entity. Dynamic signature verification has been utilized considering such scenarios. We present a new approach, namely probabilistic dynamic time warping, to verify dynamic signatures where we use dynamic time warping in realizing distance determination in the verification process. Signatures are segmented into several segments, where probability of each segment is quantified with the aid of a relative distance associated with two selected threshold levels. The final decision is achieved by combining all segment probabilities using a Bayes rule. Experiments demonstrate improvement of equal error rate for the proposed approach for the random forgery. The method has been tested on synthetic dataset and two publicly available databases of dynamic signatures, namely SCV2004 and MCYT100.

A Dempster–Shafer theory based classifier combination for online Signature recognition and verification systems

Abstract: With the advancement in technology, the society demands a robust method for person authentication. Traditional authentication methods are based on the person’s knowledge such as PIN, passwords, and tokens etc. However, such methods are prone to steal and forgotten risks. Therefore, an efficient method for person identification and verification is required. In this paper, we present a novel biometric approach for online handwritten signature recognition and verification using Dempster–Shafer theory (DST). DST has been used effectively for combination of different information sources which provide incomplete, and complementary knowledge. Initially, signature identification and verification processes have been carried out using two different classifiers, namely, Hidden Markov Model (HMM) and Support Vector Machine (SVM). Next, the performance in terms of accuracy and the reliability of the system has been increased using DST by combining the probabilistic outputs of SVM and HMM classifiers. The feasibility of the approach has been tested on MCYT DB1 and SVC2004 biometric public databases for Latin script and a new online signature dataset for Devanagari script. To our knowledge there exist no dataset on online signature available in Devanagari script. Experimental results shows that the present approach is efficient in recognition and verification of signatures and outstrips existing work in this regard till date.

In-air hand gesture signature recognition system based on 3-dimensional imagery

Abstract: A traditional online handwritten signature recognition system requires direct contact to acquisition device and usually will leave a traceable print on the surface. This made a signature possible and vulnerable to certain attempts of tracking and imitated. Looking into this shortfall, this paper proposes a novel approach to recognise an individual based on his/ her in-air hand motion while signing his/her signature. In this study, a low-cost acquisition device – Microsoft Kinect sensor is adopted to capture an image sequence of hand gesture signature. Palm region is first located and segmented through a predictive palm segmentation algorithm, which are then combined to generate a volume data. The volume data is condensed and reduced into a motion representation image by means of Motion History Image (MHI), which produces rich motion and temporal information. Several features are extracted from the MHI for empirical evaluation. Two classical recognition modes – identification and verification, are testified with an in-house database (HGS database). The proposed system achieves 90.4% identification accuracy and 3.22% equal error rate in verification mode. The experimental results substantiated the potential of the proposed system.

A verification of periodogram technique for harmonic source diagnostic analytic by using logistic regression

Abstract: A harmonic source diagnostic analytic is vital to identify the root causes and type of harmonic source in power system. This paper introduces a verification of periodogram technique to diagnose harmonic sources by using logistic regression classifier. A periodogram gives a correct and accurate classification of harmonic signals. Signature recognition pattern is used to distinguish the harmonic sources accurately by obtaining the distribution of harmonic and interharmonic components and the harmonic contribution changes. This is achieved by using the significant signature recognition of harmonic producing load obtained from the harmonic contribution changes. To verify the performance of the propose method, a logistic regression classifier will analyse the result and give the accuracy and positive rate percentage of the propose method. The adequacy of the proposed methodology is tested and verified on distribution system for several rectifier and inverter-based loads.

New Evolutionary-Based Techniques for Image Registration

Abstract: The work reported in this paper aims at the development of evolutionary algorithms to register images for signature recognition purposes. We propose and develop several registration methods in order to obtain accurate and fast algorithms. First, we introduce two variants of the firefly method that proved to have excellent accuracy and fair run times. In order to speed up the computation, we propose two variants of Accelerated Particle Swarm Optimization (APSO) method. The resulted algorithms are significantly faster than the firefly-based ones, but the recognition rates are a little bit lower. In order to find a trade-off between the recognition rate and the computational complexity of the algorithms, we developed a hybrid method that combines the ability of auto-adaptive Evolution Strategies (ES) search to discover a global optimum solution with the strong quick convergence ability of APSO. The accuracy and the efficiency of the resulted algorithms have been experimentally proved by conducting a long series of tests on various pairs of signature images. The comparative analysis concerning the quality of the proposed methods together with conclusions and suggestions for further developments are provided in the final part of the paper.

Extreme Learning Machines for Signature Verification

Abstract: In this paper, we present a novel approach to the verification of users through their own handwritten static signatures using the extreme learning machine (ELM) methodology. Our work uses the features extracted from the last fully connected layer of a deep learning pre-trained model to train our classifier. The final model classifies independent users by ranking them in a top list. In the proposed implementation, the training set can be extended easily to new users without the need for training the model every time from scratch. We have tested the state of the art deep neural networks for signature recognition on the largest available dataset and we have obtained an accuracy on average in the top 10 of more than 90%.

Implementing Signature Recognition System as SaaS on Microsoft Azure Cloud

Abstract: The use of information technology in varied applications is growing exponentially which also makes the security of data a vital part of it. Authentication plays an imperative role in the field of information security. In this study, biometrics is used for authentication purpose and also describes the combinational power of biometrics and cloud computing technologies that exhibit the outstanding properties of flexibility, scalability, and reduced overhead costs, in order to reduce the cost of the biometric system requirements. The massive computational power and unlimited storage provided by cloud vendors make the system fast. The purpose of this research is to precisely design a biometric-based cloud architecture for online signature recognition on Windows Tablet PC, which will make the signature recognition system (SRS) more scalable, pluggable, and faster, thereby categorizing it under “Bring Your Own Device” category. For extracting the features of the signature to uniquely identify the user, Webber local descriptor (WLD) process is used. The real-time implementation of this feature extraction process as well as the execution of the classifier for the verification process is deployed on Microsoft Azure public cloud. For performance evaluation, total acceptance ratio (TAR) and total rejection ratio (TTR) are used. The proposed online signature system gives 78.10% PI (performance index) and 0.16 SPI (security performance index).

Writer Independent Offline Signature Recognition Using Ensemble Learning.

Abstract: The area of Handwritten Signature Verification has been broadly researched in the last decades, but remains an open research problem. In offline (static) signature verification, the dynamic information of the signature writing process is lost, and it is difficult to design good feature extractors that can distinguish genuine signatures and skilled forgeries. This verification task is even harder in writer independent scenarios which is undeniably fiscal for realistic cases. In this paper, we have proposed an Ensemble model for offline writer, independent signature verification task with Deep learning. We have used two CNNs for feature extraction, after that RGBT for classification & Stacking to generate final prediction vector. We have done extensive experiments on various datasets from various sources to maintain a variance in the dataset. We have achieved the state of the art performance on various datasets.

VerSig: a new approach for online signature verification

Abstract: This paper introduces, VerSig, a new proposed scheme for online signature verification. The proposed scheme is based on creation of a signature envelope by employing dynamic time warping method. This envelope provides the basis for decision of forged and authentic signatures. The scheme only uses basic features such as X, Y coordinates of the signature. A well known and standardized Japanese handwritten dataset (provided for ICDAR 2013 signature verification competition) is used to evaluate the performance of proposed method. Proposed method is compared with state of art methods and observed to offer significant improvements in terms of overall accuracy of prediction.

An approach for signature recognition using contours based technique

Abstract: This research work addresses the problem of signature verification and recognition using contours based features and ANN classifier. In this work new features are proposed for signature verification and recognition. As extensive research works has been carried out in the area of HSVR signature verification and recognition by several researchers over the past two decades, many methodologies found in the literature survey. Signature plays a vital role in many fields, generally it is used for personal authentication or gaining control over a system or computing facility, physical entry to protected areas. The signature biometric problem has 2 different perspectives baed on PR. 1) Verification and 2) Recognition. In verification, features of a test signature are contrasted with features of a limited set signatures, where the class identity is claimed, whereas, in recognition, the presence of an identity test signature in the database is ascertained. There are two types of variations in signature samples Intraclass variations Interclass variations Intraclass variations are intricate variations of a person’s signature, and interclass variations are refer to variations of different person’s signature. In this exploration work, problem of effective HSVR is addressed. We propose effective feature extraction and combination of features for verification and recognition of signatures. Recognition involves FM-stage that extends to entire database.

seal or signature recognition system based on a block chain

Abstract: The invention discloses a seal or signature recognition system based on a block chain. A seal or signature provider, a seal or signature querier and a seal or signature authenticator are arranged on the block chain. Wherein the block chain data assets comprise two intelligent contracts of stamp or signature conversion and storage, stamp or signature query and comparison, and a stamp or signature database; The main function of the system is to complete collection and storage of stamps and signatures and query and identification of the stamps. The method has the advantages that the problems thatwhen people identify seals or signatures, efficiency is low, errors are likely to happen, and the seals and the signatures are effectively prevented from being falsely used in reality are solved; Through the seal or signature recognition system based on the block chain, the authenticity of the seal or signature can be efficiently and accurately recognized, and a safe and reliable operation environment is provided for electronic commerce.

Off-Line Handwritten Signature Recognition Based on Discrete Curvelet Transform

Abstract: In order to improve the offline handwritten signature recognition effect, an offline handwritten signature recognition method based on discrete curvelet transform is proposed. First, the necessary pre-processing of offline handwritten signatures is carried out, including grayscale, binarization, smooth denoising, etc. The pre-processed signature image is subjected to curvelet transform to obtain real-numbered curve coefficients in the cell matrix, and a total of 82-dimensional energy features are extracted, and multi-scale block local binary mode (MBLBP) is combined on the cell matrix of discrete curvelet transform to form a new signature feature, use the SVM classifier for training and classification. Experiments on two databases, Uyghur and Kirgiz, the highest accuracy was 97.95% and 97.42% respectively. The experimental results show that the proposed method has better accuracy in offline handwritten signature recognition.

ANN Based Handwritten Signature Recognition System

Abstract: Copyright © 2019 by author(s) and International Journal of Trend in Scientific Research and Development Journal. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0) (http://creativecommons.org/licenses/by /4.0) ABSTRACT Handwritten Signature Verification (HSV) is an automated method of verifying a signature by capturing features about a signature’s shape (i.e., static features) and the characteristics of how the person signs his/her name in realtime (i.e., dynamic features). This system provides a method of handwritten signature recognition and verification using the shapes of the signatures, artificial neural network and neural network simulation tool. The shapes of signatures are used to find the features points for features extraction. Then the extracted features are trained by using artificial neural network. A comparison of extracted features is done between the original signature and other relative signatures by using neural simulation toolbox. If the features are matched, the system shows that the signature is verified and the person is authorized and unauthorized.

Signature Recognition Based on Discrete Wavelet Transform

Abstract: Personal identification is an actively developing area of research. Human signature is a vital biometric attribute which can be used to authenticate human identity. There are many approaches to recognize signature with a lot of researches. The aim of this research is to introduce an efficient approach for signature recognition. This approach starts with the process the acquired signatures and stores these signatures in the database to be ready for verification. The collection of signature data based on collecting samples of 10 people and 10 signatures for each person through traditional ink stamp method. These signatures are digitized to be ready for processing. Many steps are applied to the acquired images to perform the pre-processing stage. The proposed approach based on discrete wavelet transforms to extract significant features from each signature image. Pre-processing is applied at the beginning of this approach to avoid any unwanted noise. This approach consists of many steps: Data acquisition, pre-processing, signature registration, and feature extraction. High recognition rate results (100%) are obtained through applying this approach.

Off-Line Writer Verification Using Segments of Handwritten Samples and SVM

Abstract: This works presents a method to verify a person identity based on off-line handwritten strokes analysis. Its main contribution is that the descriptors are obtained from the constitutive segments of each grapheme, in contrast with the complexity of the handwritten images used in signature recognition system or even with the graphemes themselves. In this way, only few handwriting samples taken from a short text could be enough to identify the writer. The descriptor is based on an estimation of the pressure of the stroke grayscale image. In particular, the average of the gray levels on the perpendicular line to the skeleton is used. A semi-automatic procedure is used to extract the segments from scanned images. The repository consists of 3.000 images of 6 different segments. Binary-output Support Vector Machine classifiers are used. Two types of cross validation, K-fold and Leave-one-out, are implemented to objectively evaluate the descriptor performance. The results are encouraging. A hit rate of 98% in identity verification is obtained for the 6 segments studied.

Neural Iris Signature Recognition (NISR)

Abstract: The security and the proper identification of individuals are vital requirements for many different applications. Biometric systems in general provide automatic recognition and identification of individuals taking advantage of the unique features of every individual. Iris recognition has a great advantage over other biometric recognition techniques, due to its huge variability of patterns among individuals. Consequently, Iris recognition tasks, even on a large database like the Chinese Academy of Sciences’ Institute of Automation (CASIA) can be searched without finding a false match. The objective of this research is to create an iris recognition system with high accuracy. This is achieved by utilizing Daugman algorithm and other techniques.

Signature Recognition and Verification Using Zonewise Statistical Features

Abstract: The signature is defined as special kind of handwriting that includes special characters and flourishes. Handwritten signatures are most accepted individual attribute for individuality authentication of the person. This provides novel method for the signature recognition and verification by using zone based statistical features. It contains mainly two phases. During the first phase, the knowledge base is constructed by training samples using the zone wise statistical features. During second stage i.e., testing phase, the processed image is obtained having zoning wise statistical features and signature is recognized using neural network classifiers. An accuracy rate of 97.5% is achieved by testing 200 samples. MATLAB is used for designing this signature recognition and verification system and is robust and avoids noise, blur and change in size, lightening conditions and other possible degradation.

A Linear-complexity Multi-biometric Forensic Document Analysis System, by Fusing the Stylome and Signature Modalities.

Abstract: Forensic Document Analysis (FDA) addresses the problem of finding the authorship of a given document. Identification of the document writer via a number of its modalities (e.g. handwriting, signature, linguistic writing style (i.e. stylome), etc.) has been studied in the FDA state-of-the-art. But, no research is conducted on the fusion of stylome and signature modalities. In this paper, we propose such a bimodal FDA system (which has vast applications in judicial, police-related, and historical documents analysis) with a focus on time-complexity. The proposed bimodal system can be trained and tested with linear time complexity. For this purpose, we first revisit Multinomial Na\"ive Bayes (MNB), as the best state-of-the-art linear-complexity authorship attribution system and, then, prove its superior accuracy to the well-known linear-complexity classifiers in the state-of-the-art. Then, we propose a fuzzy version of MNB for being fused with a state-of-the-art well-known linear-complexity fuzzy signature recognition system. For the evaluation purposes, we construct a chimeric dataset, composed of signatures and textual contents of different letters. Despite its linear-complexity, the proposed multi-biometric system is proven to meaningfully improve its state-of-the-art unimodal counterparts, regarding the accuracy, F-Score, Detection Error Trade-off (DET), Cumulative Match Characteristics (CMC), and Match Score Histograms (MSH) evaluation metrics.

Algorithm of a handwritten signature recognition online

Abstract: This document presents a technique for an online handwritten signature recognition system. A handwritten signature is an image obtained from an acquisition tool. It features dynamic and static: the dynamics are obtained from the acquisition and the static after image processing. The architecture of an online handwritten recognition system is divided into two modules, including the learning module and the module for comparison. The learning module is the creation of a model from the extracted features whereas comparison module consists in comparing the signatures to be verified or authenticated with the database models. An assessment was made by getting a false acceptance rate of 0.1% and rate of rejected true of 0.2%.

handwritten signature recognition system based on a neural network

Abstract: The invention discloses a handwritten signature recognition system based on a neural network, and aims to solve the problems that the handwritten signature recognition reliability and real-time performance are ensured during identity verification at present and the like in view of the actual field. The system comprises a handwritten signature acquisition module; A CPU module; A display module; Interface chip module. The handwritten signature acquisition module is used for acquiring a large amount of actual handwritten signature data and transmitting the data to the storage module; the data inthe storage module is imported into the CPU module through the interface chip module; the CPU module adopts a built convolutional neural network algorithm model to carry out existence on imported training data; The model is stored after unsupervised multiple iterative trainings, real-time data are transmitted into the model, whether a newly input handwritten signature is signed by a signer or notcan be quickly and accurately recognized, a recognition result is displayed on the display module, and the effect of identifying the authenticity of the handwritten signature in real time is achieved.

Enhancement of Security in Biometric System

Abstract: The Biometrics field has gained much concern in the last few years as it is an effective alternative to ancient authentication systems like passwords etc. Authentication system is needed when it is necessary to verify if a user is who they claim to be. Due to increase in traits as a result of enhancement in number of users which in turn affect the performance of database and authentication systems. The biometric systems based on single trait are on demand due to its ease of use and expedited process to access the biometric features of individuals. However, due to advent of undeniable safety factors, biometric systems are more prone to these threats. Eventually, it is the need of the hour to develop multimodal security systems which can provide highest degree security along with full accuracy and efficiency. In this Research paper a prototype for the same has been proposed. The primary focus is to design a multimodal Biometric system based on teeth, speech and signature recognition. Moreover, the highest value achieved will be treated as perfect match.