Biometric Authentication System Research Articles

Deep learning framework for biometric authentication using retinal images

ABSTRACT The inimitable blood vessel pattern of retina is a biometric feature used in person authentication. Rotation, translation and scaling of the retinal image affect the reliability and security of the biometric authentication system. Several feature extraction approaches have been proposed but to make authentication systems invariant to these variations is still challenging. This paper reports a robust retina identification system with the noble use of a convolution neural network (CNN) for automatic feature extraction. Colour image of retina is directly fed into the CNN model to extract features which are not only invariant in different geometric scales but also for illumination changes and pathological images. The network is trained using an optimisation algorithm and performed authentication using a softmax function in the final layer of the CNN model. Different CNN models are designed and tested using three groups of images consisting of coloured, grey and diseased retinal images obtained from different databases.

Design of Multiple Share Creation with Optimal Signcryption based Secure Biometric Authentication System for Cloud Environment

Biometric authentication plays a vital role in the cloud computing (CC) environment to accomplish security and privacy. Several kinds of biometrics are used for the authentication process, particularly, the fingerprint is a commonly used biometric, which accomplishes maximum security. But it can be easily stolen, unintentionally revealed, or difficult to remember. Therefore, biometric authentication techniques are needed and an encryption process is included to improve the security of the authentication process. In this aspect, this paper proposes a multiple share creation with an optimal signcryption-based biometric authentication system (MSCOS-BAS- FLNN- PeSeOA) for the cloud environment. The proposed model enables the creation of multiple shares of the input biometric image (fingerprint) and encrypts them for a secure authentication system. In addition, the encryption of individual shares takes place with an optimal signcryption system where the key generation process is carried out using the penguin search optimization algorithm (PeSeOA). Finally, the functional link neural network (FLNN) model is applied for the biometric authentication verification process. A wide range of simulation analyses are performed to highlight the supreme efficacy of the MSCOS-BAS- FLNN- PeSeOA technique. The experimental outcomes pointed out the hopeful performance of the MSCOS-BAS- FLNN- PeSeOA technique over the recent techniques.

A secure finger vein verification and authentication scheme for banking network

SummaryThe development of electronic exchanges has created a more important interest in fast and accurate customer IDs and verification. Structures, bank records, and PC framework often use personally identifiable access code IDs and test numbers (PINs) for reliable status. Biometric verification requires a level of security, unchanged structure, spoofing, and reliability. In this research work, a biometric authentication system was developed to provide security for access to the ATM network. Honest technology used for security purposes to provide efficient ATM network security is a finger vein. This method undergoes image processing technology such as preprocessing, feature extraction, and classification to locate the finger vein region for accessing the ATM network. Finally, the cascade neural network acts as a classifier method to differentiate registered and unregistered user. The performance analysis of this method is simulated by MATLAB and the process is evaluated to measure the efficient security of the network. The dataset is segregated into two stages; one for training and other for testing. The features extracted with the help of hidden layers with neurons. The matched finger vein is compared to check the accuracy.

An Improved Approach of Iris Biometric Authentication Performance and Security with Cryptography and Error Correction Codes

One of the most challenging parts of integrating biometrics and cryptography is the intra variation in acquired identifiers between the same users. Due to noise in the environment or different devices, features of the iris may differ when it is acquired at different time periods. This research focuses on improving the performance of iris biometric authentication and encrypting the binary code generated from the acquired identifiers. The proposed biometric authentication system incorporates the concepts of non-repudiation and privacy. These concepts are critical to the success of a biometric authentication system. Iris was chosen as the biometric identifier due to its characteristics of high accuracy and the permanent presence throughout an individual’s lifetime. This study seeks to find a method of reducing the noise and error associated with the nature of dissimilarity acquired by each biometric acquisition. In our method, we used Reed Solomon error-correction codes to reduce dissimilarities and noise in iris data. The code is a block-based error correcting code that can be easily decoded and has excellent burst correction capabilities. Two different distance metric measurement functions were used to measure the accuracy of the iris pattern matching identification process, which are Hamming distance and weighted Euclidean distance. The experiments were conducted with the CASIA 1.0 iris database. The results showed that the False Acceptance Rate is 0%, the False Rejection Rate is 1.54%, and the Total Success Rate is 98.46%. The proposed approach appears to be more secure, as it is able to provide a low rate of false rejections and false acceptances.

EDITH : ECG Biometrics Aided by Deep Learning for Reliable Individual Authentication

In recent years, physiological signal based authentication has shown great promises,for its inherent robustness against forgery. Electrocardiogram (ECG) signal, being the most widely studied biosignal, has also received the highest level of attention in this regard. It has been proven with numerous studies that by analyzing ECG signals from different persons, it is possible to identify them, with acceptable accuracy. In this work, we present, EDITH, a deep learning-based framework for ECG biometrics authentication system. Moreover, we hypothesize and demonstrate that Siamese architectures can be used over typical distance metrics for improved performance. We have evaluated EDITH using 4 commonly used datasets and outperformed the prior works using less number of beats. EDITH performs competitively using just a single heartbeat (96-99.75% accuracy) and can be further enhanced by fusing multiple beats (100% accuracy from 3 to 6 beats). Furthermore, the proposed Siamese architecture manages to reduce the identity verification Equal Error Rate (EER) to 1.29%. A limited case study of EDITH with real-world experimental data also suggests its potential as a practical authentication system.

SIFT Enhanced Multiple Biometric Authentication System Integrated With Temperature Sensor

Biometric authentication is a security procedure used to verify a person's identity, based on his or her unique physical characteristics. It is used to control access to both physical and digital resources, such as organizations, rooms, and different gadgets. Multiple biometric authentication systems use more than one physiological feature of an individual for the purpose of identification and verification and they are a superior solution for every industry that demands higher accuracy as well as security. This method proposes a contact less multiple biometric authentication system which uses palm print and hand geometry features. A temperature sensor is integrated in this system for detecting elevated body temperatures. A high body temperature or fever is related to many diseases and can be a first indicator whether a person is healthy or not. Using the SIFT (Scale Invariant Feature Transform) algorithm, this design achieves robust local feature detection and extraction in an image.

Construction and Application of Automatic Attendance Prediction System for Classroom Education Teaching Based on Random Matrix Theory

Random matrix theory, as a theoretical tool for statistical processing of complex systems and high-dimensional data, has been more and more combined with big data technology and applied to automatic attendance prediction analysis of classroom education. Firstly, the basic definition of the random matrix is introduced. Secondly, the single ring theorem of the random matrix is introduced. Based on this, the random matrix model of data is established. However, few undergraduate or graduate students in physics have been exposed to this theory. By introducing the basic concept of random matrix theory, this paper analyzes the necessity and feasibility of introducing random matrix theory into automatic attendance prediction teaching in classroom education. As a biometrics technology, random matrix theory technology has been widely used in various fields of the identity verification system, and classroom attendance system is an important field of random matrix theory technology application. The traditional random matrix theory class attendance has the problem of slow recognition speed and low accuracy. With the popularity of deep learning, the random matrix theory based on deep learning has gradually replaced the traditional random matrix theory. In this paper, the SSD target detection algorithm, NB algorithm, and RF algorithm based on deep learning are used to improve and optimize the traditional random matrix theory class attendance system, which effectively improves the efficiency and accuracy of random matrix theory attendance. Then, aiming at the low recall rate and recognition rate of random matrix theory in low-pixel images, we try to apply the random matrix theory of single image resolution reconstruction method to face detection and random matrix theory and give a calculation method to deal with outliers and noise.

Availability and Use of Information and Communication Technologies at an International Airport in Nigeria

In this paper, the availability and use of conventional and aviation-specific information and communication technologies by airport workers are examined. A sample survey research design and a purposive sampling technique were adopted. By using a questionnaire, quantitative data were collected from 213 staff of the airport, who were available during the survey and were willing to complete the instruments. Further, qualitative data were collected from two heads of departments by using an interview schedule. The conventional technologies of computers, printers, and photocopiers, among others, were reportedly available and used. Common-use terminal equipment and flight information display systems were reportedly the aviation-specific technologies known by the largest number of respondents (84% each), while the biometric identification system was known by the least number of respondents (43.6%). The respondents reported increasing use of these technologies in the last five years. The members of staff of the organisation have very fair opinions about their knowledge of the use and deployment of information and communication technologies, but they identified infrastructural constraints and the high cost of equipment maintenance as hitches to the adequate deployment of the ICTs. Continuous training—that covers both conventional ICT and aviation-specific information and communication technologies—is required to improve the performance of airport staff.

Stable Hash Generation for Efficient Privacy-Preserving Face Identification

The development of large-scale facial identification systems that provide privacy protection of the enrolled subjects represents an open challenge. In the context of privacy protection, several template protection schemes have been proposed in the past. However, these schemes appear to be unsuitable for indexing (workload reduction) in biometric identification systems. More precisely, they have been utilized in identification systems performing exhaustive searches, thereby leading to degradations of the computational efficiency. In this work, we propose a privacy-preserving face identification system which utilisers a Product Quantization-based hash look-up table for indexing and retrieval of protected face templates. These face templates are protected through fully homomorphic encryption schemes, thereby guaranteeing high privacy protection of the enrolled subjects. For the best configuration, the experimental evaluation carried out over closed-set and open-set settings shows the feasibility of the proposed technique for the use in large-scale facial identification systems: a workload reduction down to 0.1% of a baseline approach performing an exhaustive search is achieved together with a low pre-selection error rate of less than 1%. In terms of biometric performance, a False Negative Identification Rate (FNIR) in range of 0.0% - 0.2% is obtained for practical False Positive Identification Rate (FPIR) values on the FEI and FERET face databases. In addition, our proposal shows competitive performance on unconstrained databases, e.g., the LFW face database. To the best of the authors’ knowledge, this is the first work presenting a competitive privacy-preserving workload reduction scheme which performs template comparisons in the encrypted domain.

"Iris Recognition Based Data Security for Secure Transmission Over Internet of Things Network"

The biometric traits of any human being such as fingerprint, tongue prints, face recognition, Retina Scan, Iris recognition etc. are unique and cannot be replicated or fabricated using modern technologically advanced processes. These unique traits are used to identify individuals for grating access to sensitive data over the cloud. In this paper we propose a biometric authentication system based on Iris recognition to allow user identity based data to be pushed on to the cloud for global access, if the user Iris image is authenticated with the database. The real time Iris scanner captures the Iris image and same is imported to the MATLAB workspace for feature extraction and matching. Upon successful matching, the data is transmitted to the ThingSpeak Internet of Things (IoT) analytics platform. The IoT framework is a massive network of physical devices sharing data according to set protocols. Most of the IoT enabled devices lack security protocols. Thus, biometric authentication in general and iris recognition in particular finds a perfect fit for securing data over an IoT framework.

EEG-based Biometric Authentication Using Machine Learning: A Comprehensive Survey

An electroencephalogram (EEG) is a measurement that reflects the overall electrical activity in the brain. EEG signals are effective for biometric authentication and robust against malware attacks and any kind of fraud activities due to the uniqueness of the signals. Significant progress in research on EEG-based authentication has been achieved in the last few years, with machine learning being extensively used for classifying EEG signals. However, to the best of our knowledge, there has been no investigation into the overall progress made in such research. In this paper, the literature on the various factors involved in state-of-the-art biometric authentication systems is reviewed. We provide a thorough comparison of different machine learning biometric authentication techniques. The comparison criteria include the research objectives, machine learning algorithms, computational complexity, source of brainwaves, feature extraction methods, number of channels, and so on. Alongside the discussion of existing works, directions for future research are suggested to improve authentication accuracy. This paper provides an in-depth discussion of different advanced biometric authentication techniques, and a vivid picture of state-of-the-art machine learning-based biometric authentication techniques using EEG.

ПОВЫШЕНИЕ ФУНКЦИОНАЛЬНОЙ НАДЕЖНОСТИ СИСТЕМ БИОМЕТРИЧЕСКОЙ ИДЕНТИФИКАЦИИ ПУТЕМ ВНЕСЕНИЯ ИЗБЫТОЧНОСТИ

The paper demonstrates a need to study the functional reliability of identification systems and the applicability of methods for ensuring the functional reliability of information systems for biometric identification systems. Methods of ensuring functional reliability of biometric identification systems via providing indicators of reliability of identification devices and introducing redundancy are considered. The applicability of the methods for biometric identification systems is investigated, its advantages and disadvantages are evaluated. Triple modular redundancy scheme has been developed for the program redundancy using N-version programming. An appropriate reliability scheme based on the principles of dual programming has been developed to implement the identification system with one additional redundant element. Based on the NIST statistics of face recognition algorithms for the developed reliability schemes the reliability indicators were evaluated and its advantages over a system consisting of a single non-redundant element have been shown. The proposed reliability schemes are consistent with the practice of ensuring the functional reliability of information systems and can be integrated into biometric identification systems that use biometric data as the main access identifier.

Personal Authentication System Based on Iris Recognition and Digital Signature Technology

Authentication based on biometrics is being used to prevent physical access to high-security institutions. Recently, due to the rapid rise of information system technologies, biometrics is now being used in applications to access databases and commercial workflow systems. These applications need to implement measures to counter security threats. Many developers are exploring and developing novel authentication techniques to prevent these attacks. However, the most challenging problem is how to keep biometric data while maintaining the functional performance of identity verification systems. This paper presents a biometrics-based personal authentication system in which a smart card, a Public Key Infrastructure (PKI), and iris verification technologies are combined. Raspberry Pi 4 Model B+ is used as the core of hardware components with an IR Camera. Following that idea, we designed an optimal image processing algorithm in OpenCV, Python, Keras, and sci-kit learn libraries for feature extraction and recognition is chosen for application development in this project. After training, the implemented system gives an accuracy of (97% and 100%) for the left and right (NTU) iris datasets. Later, the person verification based on the iris feature is performed to verify the claimed identity and examine the system authentication. For the NTU iris dataset, the time of key generation, Signature, and Verification is 5.17sec,0.288, and 0.056, respectively. This work offers the realistic architecture to implement identity-based cryptography with biometrics using the Rivest–Shamir–Adleman (RSA) algorithm.

МЕТОДЫ УСИЛЕНИЯ ПРОЦЕДУРЫ ИДЕНТИФИКАЦИИ ПОЛЬЗОВАТЕЛЕЙ НА ОСНОВЕ ТЕХНОЛОГИИ LIVENESS DETECTION

Biometric identification and access control systems contain methods for recognizing a subject'spersonality based on his unique physiological and behavioral characteristics. The purpose ofthis work is to develop a system for secure interaction (authentication) of participants in gamifiededucational projects, which includes countering security threats that arise when using biometric usercharacteristics. A comparative analysis of the efficiency of recognition of fake biometric samples byliveness detection methods based on the detection of sample substitution using a photo, video on adisplay, a 3D model, and a mask has been performed. During research a method of using the livenessdetection for include to a gamified educational environment system was proposed. A modification ofthe liveness detection method (hybrid method) has been proposed and a biometric identification systemin real time has been designed using the proposed method. A two-stage hybrid biometric identificationmethod has been developed based on the joint use of passive and active software methods fordetecting fake biometric samples. The method is adapted for use with a minimum number of additionaldevices, the only biometric feature scanner is a 2D-camera. The network of types two-layerperceptron, three-layer perceptron and convolutional neural network was tested. The network wastrained on the author's training examples. The position of the announcer when recording trainingexamples: the distance of the face from the camera is 60cm, the recording modes when the head isturned by 0 (look directly into the camera), 30 (the head is slightly turned to the side) and 45 (thehead is turned strongly to the side) degrees. Based on the testing results, the best recognition rateswere found in a convolutional neural network with 3 convolutional layers and 1 fully connected one.Accuracy of recognition of the spoken word is obtained up to 100% when the user's head is turned upto 30° and up to 70% - when the user's head is turned up to 45°. The FAR value of this system was1%, the FRR value was 0% for testing on 1000 samples.

Performance analysis of a parameterized minutiae-based approach for securing fingerprint templates in biometric authentication systems

The fingerprint templates include almost the entire information about an individual’s fingerprints, and the user validation is totally determined by the fingerprint minutiae features. Due to the sort of attacks, the most challenging task is to shield the fingerprint templates (or minutiae features). The paper is concerned with the development of a parameterized minutiae-based approach to generate the secure biometric templates from the fingerprint images. The minutiae features were extracted from the individual’s fingerprint images, and then their associated parameters were explored to hide the real minutiae information. A minutiae-based transformation algorithm was employed by adding extra chaffs (fake minutiae) and remodeling the real minutiae to generate a secure vault that leads to a cancelable fingerprint template. As a fingerprint template contains only the transformed and fake minutiae, it is almost impossible for the attacker to induce the real minutia information. The proposed method assisted to create multiple dissimilar fuzzy vaults for a single fingerprint; thereby, the correlation attacks would be resisted. The proposed method achieved the highest accuracy in verification phase compared to other methods, and provided a novel contribution in the field of biometric security for protecting the fingerprint templates.

An ArmurMimus multimodal biometric system for Khosher authentication

AbstractOne of the basic requirements of our modern day society is personal authentication. Biometric recognition should make a human‐like identity determination by identifying its physiological and/or behavioral characteristics. In comparison to traditional knowledge‐based approaches, biometric identification systems have the potential to bring benefits. However, because of the difficulties in extracting non‐class discriminative features, the lack of protection during storage of extracted features, and poor recognition accuracy, most frequently used biometric systems lack model protection and robustness. This research proposed a Mimus multimodal biometric system focused on the combination of multiple modalities and optimal level fusion of features to resolve these problems. Initially, the novel Blob‐funk method extracts the complementary non‐class discriminatory information among different modalities, which accomplishes the biometric data enrollment. Thus, it extracts the different properties by comparing surrounding regions based on finding the local maxima and minima of the function. After extracting the features, they need to be stored in a secure manner in a database. Therefore, the paper incorporates the new code block protection strategy to achieve an effectual protection of continuous monitoring via the generation of non‐invertible features, which is used to create the templates, thus storing them in a database. Finally, the novel Lucynomial logistic regression system incorporates user authentication and thus achieves greater recognition accuracy through estimation of threshold value with confrontation of spoof attacks. Hence, compared to the existing techniques such as SVM, PCA, and DBN, the outcome of the proposed work attains 97.53% accuracy, 0.020% FAR, 96.44% recall, and 97.85% precision, thus exemplifying the competence of the novel system.

Combining Iris, Sclera and Pupil Features for Biometric Authentication System on Smartphone Devices

Currently, biometric authentication systems are commonly used based on physical and behavioural biometric modalities like iris, face, fingerprints, ear, sclera, DNA, voice, signature, etc. Rather than relying on the standalone or unimodal biometric system, multimodal biometric systems are secure and provide more accurate results for person identification and verification. This paper introduces the multimodal eye biometric authentication system where iris, pupil and sclera features are extracted using CNN based on entropy values to perform the accurate automatic segmentation for smartphone devices. The eye images used in the proposed approach for training and testing are completely captured by smartphones. The fusion method used to fuse the colour and texture characteristics of iris and pupil with Y-shaped sclera characteristics from eye image based on support value is Feature Level Fusion. As the images are captured in normal environment settings, it is an unconstrained colour eye image database. MATLAB is used for the experimentation and testing of the model. The proposed eye biometric system outperforms in the case of segmentation and recognition accuracy. Recognition accuracy is –% for unconstrained eye images achieved for the eye image database captured by smartphones.

ECG classification system based on multi-domain features approach coupled with least square support vector machine (LS-SVM)

Developing a robust authentication and identification method becomes an urgent demand to protect the integrity of devices data. Although the use of passwords provides an acceptable control and authentication, it has shown much weakness in terms of speed and integrity, which make biometrics the ideal authentication solution. As a result, electrocardiogram (ECG) signals have received a great attention in most authentication systems due to the individualized nature of the ECG signals that make them difficult to counterfeit and ubiquitous. In this paper, we propose a new model for ECG verification using multi-domain features coupled with a least square support vector machine (LS-SVM). Two types of features are investigated to find the best set of features to individual from ECG signals. Time domain and frequency domain features based on optimized Triple Band filter bank are extracted from ECG signals. The extracted features are investigated to figure out the best relevant features and remove the redundant ones. The selected features are fed to three classifiers, including Least Square Support Vector Machine (LS-SVM), K-means, and K-nearest. The obtained results have shown that our ECG biometric authentication system outperforms existing methods. The proposed model obtained an average of accuracy of 88%, 95% with time and frequency features, respectively, while it recorded 99% when a combination of time and frequency features are used to classify ECG signals. A public dataset is used to assess the proposed model.

Biometric Identification System Based on Contactless Palm-Vein Using Residual Attention Network

Palm vein recognition technology is a one of the most effective biometric technologies for personal identification. Palm acquisition techniques are either contact-based or contactless-based. The contactless-based palm vein system is considered more accurate and efficient when used in modern applications, but it may suffer from problems like pose variations and the delay in the matching process. This paper proposes a contactless-based identification system for palm vein that involves two main steps; First, the central region of the palm is cropped using fast extract region of interest algorithm, then the features are extracted and classified using altered structure of Residual Attention Network, which is a developed version of convolutional neural network that uses an attention mechanism. The altered structure is constructed by stacking multiple attention modules and pre-activation residual unit with additional down sampling layers in between. The proposed system was tested on contactless CASIA multispectral palm vein databases that contains palm images with obvious pose variations taken from 100 persons. The results show that our system has outperformed other state-of-the-art systems with 95.55% accuracy and fast identification process of 0.06 second per person.

A high-level synthesis approach for precisely-timed, energy-efficient embedded systems

Embedded systems continue to rapidly proliferate in diverse fields, including medical devices, autonomous vehicles, and more generally, the Internet of Things (IoT). Many embedded systems require application-specific hardware components to meet precise timing requirements within limited resource (area and energy) constraints. High-level synthesis (HLS) is an increasingly popular approach for improving the productivity of designing hardware and reducing the time/cost by using high-level languages to specify computational functionality and automatically generate hardware implementations. However, current HLS methods provide limited or no support to incorporate or utilize precise timing specifications within the synthesis and optimization process. In this paper, we present a hybrid high-level synthesis (H-HLS) framework that integrates state-based high-level synthesis (SB-HLS) with performance-driven high-level synthesis (PD-HLS) methods to enable the design and optimization of application-specific embedded systems in which timing information is explicitly and precisely defined in state-based system models. We demonstrate the results achieved by this H-HLS approach using case studies including a wearable pregnancy monitoring device, an ECG-based biometric authentication system, and a synthetic system, and compare the design space exploration results using two PD-HLS tools to show how H-HLS can provide low energy and area under timing constraints. © 2021 Elsevier Science. All rights reserved