Iris Recognition Research Articles

Bias Cuts and Data Dumps

Bias Cuts and Data Dumps

FPGA-Based CNN for Eye Detection in an Iris Recognition at a Distance System

Neural networks are the state-of-the-art solution to image-processing tasks. Some of these neural networks are relatively simple, but the popular convolutional neural networks (CNNs) can consist of hundreds of layers. Unfortunately, the excellent recognition accuracy of CNNs comes at the cost of very high computational complexity, and one of the current challenges is managing the power, delay and physical size limitations of hardware solutions dedicated to accelerating their inference process. In this paper, we describe the embedding of an eye detection system on a Zynq XCZU4EV UltraScale+ multiprocessor system-on-chip (MPSoC). This eye detector is used in the application framework of a remote iris recognition system, which requires high resolution images captured at high speed as input. Given the high rate of eye regions detected per second, it is also important that the detector only provides as output images eyes that are in focus, discarding all those seriously affected by defocus blur. In this proposal, the network will be trained only with correctly focused eye images to assess whether it can differentiate this pattern from that associated with the out-of-focus eye image. Exploiting the neural network’s advantage of being able to work with multi-channel input, the inputs to the CNN will be the grey level image and a high-pass filtered version, typically used to determine whether the iris is in focus or not. The complete system synthetises other cores and implements CNN using the so-called Deep Learning Processor Unit (DPU), the intellectual property (IP) block released by AMD/Xilinx. Compared to previous hardware designs for implementing FPGA-based CNNs, the DPU IP supports extensive deep learning core functions, and developers can leverage DPUs to conveniently accelerate CNN inference. Experimental validation has been successfully addressed in a real-world scenario working with walking subjects, demonstrating that it is possible to detect only eye images that are in focus. This prototype module includes a CMOS digital image sensor that provides 16 Mpixel images, and outputs a stream of detected eyes as 640 × 480 images. The module correctly discards up to 95% of the eyes present in the input images as not being correctly focused.

Enhancing Biometric Security: A Framework for Detecting and Preventing False Identification

Biometrics is a technological system that utilizes data to differentiate one individual from another. The biometric framework can be used by government and private organizations for security purposes. This software-based technology helps to look at an individual's data if it is genuine or fake. The study suggested a framework; its goal is to strengthen the development and acceptance of the biometric system. The function of this system is to reduce the applied effort to identify and recognize the quality of the image in less time. This study utilizes three data applications: iris, fingerprint, and face recognition. The approach proposed by the survey uses different features of the images to determine the difference between the original image and the considered sample image. It gives efficient protection against different spoofing attacks. Simulation results show that the high-quality detection application has an average peak signal-to-noise ratio (PNSR) of 89.77. Further, the proposed model effectively detects false biometric identification.

Face Recognition Based Attendance System Using Histogram of Oriented Gradients and Linear Support Vector Machine

In the 21st century, modern technology is playing an important role in providing innovative on traditional challenges across various domains or sectors. One such challenging task is of daily attendance marking and tracking. Manual attendance requires efforts and it is time-consuming. Sometimes attendance cannot be mark due to human errors. Relying on voice, iris, or fingerprint recognition, increases the complexity and the hardware infrastructure of the system and also increases the cost. To effectively address such issues, we have developed a “Camera based Attendance System”. This system encompasses several crucial stages, including data entry, dataset of multiple people. It is an image-based face recognition system for marking attendance on the SQL database. It excels in detecting and recognizing multiple individuals faces from image and comparing it with the dataset for accurately marking the attendance. This makes the attendance marking process fully automatic. Remarkably, our proposed system attains an impressive recognition and provides the accuracy of approximately 95%. With this solution, daily attendance marking and recording becomes effortless and the stored attendance record can be also used in future if require, eliminating the risk of attendance not getting marked due to human error.

Iris Scan with 3 Blinks Plus Emotion Recognition for Secure Payment Method for POS using Cloud Computing

Financial transactions are becoming unsecured day by day by means of phishing, stolen identity or stolencredentials. It is the need of the hour to innovate convenient payment methods that are secure. Bearing this in mind we through this paper propose a system that is complex to developbut does achieve the targets of providing convenient and secured payment financial services. To make financial transactions we now suggest Face Recognition as the first parameter, 2nd parameter be emotions recognition and 3rdbeing 3 blinks IRIS scan. Once all 3 are positive then thetransaction will be successful.

Study the effect of eye diseases on the performance of iris segmentation and recognition using transfer deep learning methods

A new deep learning-based iris recognition system is presented in the current study in the case of eye disease. Current state of art iris segmentation is either based on traditional low accuracy algorithms or heavy-weight deep-based models. In the current study segmentation section, a new iris segmentation method based on illumination correction and a modified circular Hough transform is proposed. The current method also performs a post-processing step to minimize the false positives. Besides, a ground truth of iris images is constructed to evaluate the segmentation accuracy. Many deep learning models (GoogleNet, Inception_ResNet, XceptionNet, EfficientNet, and ResNet50) are applied through the recognition step using the transfer learning approach. In the experiment part, two eye disease-based datasets are used. 684 iris images of individuals with multiple ocular diseases from the Warsaw BioBase V1 and 1,793 iris images from the Warsaw BioBase V2 are also used. The CASIA V3 Interval Iris dataset, which contains 2,639 photographs of healthy iris, is used to train deep models once, and then the transfer learning of this normal-based eye dataset is used to retrain the same deep models using Warsaw BioBase datasets. Different scenarios for training and evaluating participants are used during experiments. The trained models are evaluated using validation accuracy, training time, TPR, FNR, PPR, FDR, and test accuracy. The best accuracies are 98.5% and 97.26%, which are recorded by the ResNet50 (2-layer of transfer learning) model trained on Warsaw BioBase V1 and V2, respectively. Results indicate that the effect of eye diseases is concentrated on the segmentation phase. For recognition, no significant impact is recognized. Some disease that affects the structure (bloody eyes, trauma, iris pigment) can affect the iris recognition step partially. Our study is compared with similar studies in the case of eye diseases. The comparison proves the efficiency and high performance of the proposed methodology against all previous models on the same iris datasets.

A cascaded deep learning framework for iris centre localization in facial image

AbstractAccurate iris centre localization is crucial in many computer vision and facial biometric applications such as gaze estimation, human–computer interaction, iris recognition, and liveness detection. However, it is challenging in an uncontrolled environment due to variations like pose, scale, rotation, specular reflection, and image quality. Therefore, a cascaded deep learning framework for iris centre localization in facial images is proposed that is robust to the abovementioned variations. The proposed approach consists of (i) YOLOv3 for eye detection, (ii) UNet for iris segmentation, and (iii) statistical modelling for iris centre localization. The eyes are first detected using the YOLOv3, and subsequently, iris segmentation is performed within the detected eyes using the UNet. Following iris segmentation, statistical modelling is employed to enhance the localization accuracy of the iris centre. Experiments were performed on benchmark databases, resulting in a standardized error measure SED of 3.405 pixels for BioID and 3.259 pixels for GI4E databases. In addition, the robustness of the proposed eye detection model was further evaluated on the Yale B for illumination variations and the CAS‐PEAL for pose variations.

Security enhancement in a cloud environment using a hybrid chaotic algorithm with multifactor verification for user authentication

A hybrid chaotic-based DNA and multifactor authentication strategy are created to improve the protection of the cloud environment. Initially, multimodal data are collected from the data owner then the information is compressed by utilizing the deflate compression approach. The data is then encrypted using hybrid chaotic-based DNA cryptography to increase the security of data. In this hybrid algorithm DNA is used for the key generation and chaotic algorithm is utilized for the encryption process. On the other hand, a multifactor authentication method is created to access data from the cloud to block access by unauthorized users. In that technique, users are requested to enter the registered Password with the generated OTP from the mobile number. Then, the device serial number is another factor to verify the accessing device. Likewise, the user's fingerprint and iris recognition are also validated for accessing the data. The cloud-based data can be accessible following users' successful authentication. The simulation analysis shows that the encryption and decryption time reached for image, string and integer data is 24, 0.065, 37, 0.14, 28 and 0.14 s, respectively, for the cloud security algorithm. The proposed algorithm effectively mitigates space consumption and provides improved data security in a cloud environment.

Design and Analysis of an Expert System for the Detection and Recognition of Criminal Faces

The process of identifying a person using their facial traits is referred to as face recognition, and it is a form of biometric identification. The use of facial recognition might range from that of an entertainment tool to one of a security tool. Even while other forms of biometric identification, such as fingerprints and iris scans, are reliable, they require the active participation of an individual. As a result, criminals cannot rely on them as the most reliable means of verification. When a criminal database, which stores the individual details of a criminal, and facial recognition technology are brought together, it can identify a criminal who is depicted in an image or seen in a video feed. Not only does a criminal recognition system needs to have a high level of accuracy, but it also needs to be able to adapt to significant changes in lighting, occlusion, aging, expressions, and other factors. In this study, they were analyzed and compared with the many methods of face detection and face recognition, such as HAAR cascades, local binary patterns histogram, support vector machines, convolutional neural networks, and ResNet-34. These methods include a variety of different approaches to recognizing faces. An analysis of these strategies is also conducted and then put into practice to those that seem to be the most effective for the designed criminal recognition system. In addition to that, a variety of uses of this criminal recognition in the real world are also discussed.

Exploiting Deep Learning Techniques for the Verification of Handwritten Signatures

Biometric featuresare common measures of identity verification where signaturesarethe most used type. The digital technology has given birth to new ways of biometric identification, such as fingerprints, iris and face recognition,while dealing with handwritten signatures is still a challenging task, because handwritten signatures are more prone to forgery than other means of verification due to issues like computer error, insufficient datasets, and loss of information. This work aims to develop a system that takes a signature image as its input and determines whether the signature is genuine written by its author or forged by another individual. The system is based on a neural network algorithm called Convolutional Siamese Neural Networks, which is used for deep learning and computer vision as well as other machine learning tasks such as natural language processing and digital signal processing.A Contrastive Loss function which compares the Euclidean distance of the output feature vectors is used, and a writer-independent model is used for training and image classification. This work’s objective is toenhance the precision of signature verification and take it as a base for future work on signature verification and use it in user identification, fraud detection and prevention, and forensic investigation applications. The system can be applied in banking, government and private organizations, and forensic investigation for identity and document verification, impersonation and fraud detection and prevention, crime and judicial investigation, and passport verification

LRFID-Net: A Local-Region-Based Fake-Iris Detection Network for Fake Iris Images Synthesized by a Generative Adversarial Network

Iris recognition is a biometric method using the pattern of the iris seated between the pupil and the sclera for recognizing people. It is widely applied in various fields owing to its high accuracy in recognition and high security. A spoof detection method for discriminating a spoof attack is essential in biometric recognition systems that include iris recognition. However, previous studies have mainly investigated spoofing attack detection methods based on printed or photographed images, video replaying, artificial eyes, and patterned contact lenses fabricated using iris images from information leakage. On the other hand, there have only been a few studies on spoof attack detection using iris images generated through a generative adversarial network (GAN), which is a method that has drawn considerable research interest with the recent development of deep learning, and the enhancement of spoof detection accuracy by the methods proposed in previous research is limited. To address this problem, the possibility of an attack on a conventional iris recognition system with spoofed iris images generated using cycle-consistent adversarial networks (CycleGAN), which was the motivation of this study, was investigated. In addition, a local region-based fake-iris detection network (LRFID-Net) was developed. It provides a novel method for discriminating fake iris images by segmenting the iris image into three regions based on the iris region. Experimental results using two open databases, the Warsaw LiveDet-Iris-2017 and the Notre Dame Contact Lens Detection LiveDet-Iris-2017 datasets, showed that the average classification error rate of spoof detection by the proposed method was 0.03% for the Warsaw dataset and 0.11% for the Notre Dame Contact Lens Detection dataset. The results confirmed that the proposed method outperformed the state-of-the-art methods.

Iris Recognition in Cross-Spectral Based on Histogram of Oriented Gradient and Linear Discriminant Analysis

The use of biological and behavioural characteristics has significantly increased in recent years. Like fingerprints, faces, iris, and others in numerous crucial applications in security and electronic governance. Where recently, Research has concentrated on illuminating the recognition of iris images obtained in different fields across spectra by sensors that can capture double iris images in various areas (environment/lighting) by infrared Nearby and visible illumination (NIR, VIS) due to features unique in the iris texture that differ for the same individual, between their left and right eyes. But there is challenging to match NIR and VIS photographs due to the spectrum differences between pictures captured by near-infrared and visible light. We suggest utilizing the Histogram of Oriented Gradient (HOG) approach to extract meaningful information from images of the iris and implement Normalization to reduce differences in illumination, also apply the LDA for dimensionality reduction and K-Nearest Neighbor and Support Vector Machine as a classifier. The experimental findings were implemented on the PolyU bi-spectral dataset, and the results show that the suggested approach achieved exceptional matching performance by KNN, SVM on (Right to Left) for each (NIR, VIS) or versa (99.14%,99.04%, 97.44%,96.38%) respectively.

Efficient Iris Image Recognition System Based on Machine Learning Approach

Iris identification is a well‐known technology used to detect striking biometric identification techniques for recognizing human beings based on physical behavior. The texture of the iris is essential and its anatomy varies from individual to individual. Humans have distinctive physical characteristics that never change. This has resulted in a considerable advancement in the field of iris identification, which inherits the random variation of the data and is often a dependable technological area. This research proposes three algorithms to examine the classifications in machine learning approaches using feature extraction for the iris image. The applied recognition system used many methods to enhance the input images for iris recognition using the Multimedia University (MMU) database. Linear Discriminant Analysis (LDA) feature extraction method is applied as an input of three algorithms of machine learning approaches that are OneR, J48, and JRip classifiers. The result indicates that the OneR classifier with LDA achieves the highest performance with 94.387 % accuracy, while J48 and JRip reached to 90.151% and 86.885% respectively. Index Terms— Iris recognition, LDA, OneR, J48, JRip.

Harnessing highly efficient triboelectric sensors and machine learning for self-powered intelligent security applications

In the contemporary epoch, distinguished by a transition from the internet-of-things (IoT) to the artificial intelligence of things (AIoT), individual electronic appliances necessitate inherent power-generation, independence from internet connectivity, and an imbued degree of intellect. Devices governed by pressure or strain sensors particularly demand such attributes. Responding to this technological imperative, our study endeavored to conceive an intelligent door security apparatus grounded on the universally adopted numerical input system. Despite the commercialization of identification systems such as fingerprint, iris, or facial recognition, these mechanisms suffer from susceptibility to a variety of functional aberrations. Consequently, our investigation concentrated on a security system predicated on numerical input. This necessitated the formulation of a swift, self-powered pressure sensor characterized by sensitivity to minute pressure changes. As such, we engineered a triboelectric pressure sensor incorporating a composite of Ti3C2-based MXene and polydimethylsiloxane (PDMS) as the electronegative stratum, and Nylon functioning as the electropositive layer. Addressing the sensor's intrinsic deficiency in sensitivity to pressure, we augmented the MXene-PDMS composite's surface with an out-of-plane wavy structure, and utilized a Nylon stratum composed of nanofibers, thereby amplifying the contact area under pressurized conditions. This meticulously developed sensor displayed a sensitivity metric of 0.604 kPa−1 at 15 kPa, and notably, the swiftest response times recorded amongst triboelectric pressure sensors to date. Post attachment of the sensor to a numeric keypad (ranging from 0 to 9), we meticulously measured the signal alterations contingent on each key press, resulting in a comprehensive dataset. Employing a multitude of machine learning algorithms, we realized an exemplary degree of precision in both training and testing phases. The pragmatic implications of this work are noteworthy. Not only does our technology facilitate the unlocking of a door by entering the correct numerical code, but it is capable of recognizing distinct triboelectric signal patterns, corresponding to the specific manner of key entry by an authorized user, offering an additional dimension of security.

Instant learning based on deep neural network with linear discriminant analysis features extraction for accurate iris recognition system

Instant learning based on deep neural network with linear discriminant analysis features extraction for accurate iris recognition system

Retina image assessment for microstructural difference detection

The dissimilarities in appearance of the iris texture patters can improve the individual’s recognition and can strength the capabilities of multibiometrics systems. There is an important within-person variation or intraclass variation that produces different patters such as the geometry or iris pigmentation details that change over time or could be different when the right iris and left iris are analyzed. Moreover, the sensing environment such as ambient lighting variations, eye rotation due to the head tilt or inconsistent iris size due to the distance from the camera can also introduce within-person variation. In this paper, we apply the fuzzy edges detection, in a comparative fashion between right iris and left iris, to find the interoperability capability of a particular biometric trait determined primarily by genotype. In other words, we investigated the degree of variation of the irises of the same person. The structural similarity index measure (SSIM) is implemented to investigate the structural similarity between two iris codes. Prior to the similarity analysis, the segmented iris (i.e., annular area between pupil and sclera) is the edges detection based on the fuzzy edge is performed. This operation allows comparisons between the right iris and left iris without any influence of the stretch or dilation of the pupil induced by different illumination conditions. Also, we can estimate if the left and right irises belong to the same or to different individuals. The proposed approach is tested on the MMU1 Iris Database (with 225 images of the left eye and 225 images of the right eye). An average SSIM value of 0.9216, indicates that proposed iris biometrics model effectively differentiates between left and right eyes of the same person. Also, this result indicates that there is recognizable similarity between left and right irises. These results could be useful for certain applications devoted to detect anomalies in the human irises that could be associated to various diseases.

Research on Iris Feature Extraction and Recognition Technology Based on Deep Learning

Abstract Certain biological information or behavioral information of a person can achieve the effect of characterizing an individual, and by combining the computer to extract the corresponding information, identity authentication is achieved. In a variety of biometrics, iris relative to fingerprints, handwriting and face, belongs to the structure within the human eye, if you want to steal is very difficult, in order to improve the safety factor, so the iris is used for authentication to achieve iris recognition. This study is based on deep learning iris recognition matching, in order to be able to effectively improve the accuracy of iris recognition, experiments are carried out. Evaluation metrics are performed through Hamming distance to calculate the correct recognition rate to ensure that the iris information can be accurately represented. This study mainly uses the improved PCHIP-LMD algorithm and CNN algorithm, the LMD algorithm is more context-aware, has better generalization ability, flexibility and scalability, while the CNN algorithm has the advantages of local awareness, parameter sharing and automatic parameter learning. In this study, we compare the correct recognition rate of iris recognition between improved PCHIP-LMD and CNN algorithms and get the conclusion that the correct recognition rate of the improved PCHIP-LMD algorithm is only 78%, which is much smaller than that of the CNN algorithm which is 92%, and we get the conclusion that LMD algorithm is suitable for iris recognition with few samples, and it is more suitable to use CNN algorithm when the sample images are too many. CNN algorithm. With the development of technology, the application of iris recognition will be more and more, I believe that soon will be widely popularized in daily life and work.

The discrete wavelet transform based iris recognition for eyes with non-cosmetic contact lens

Iris recognition has been used as one of the biometric systems for user authentication, identification, and verification for quite some time. The basis of an iris recognition lies on the matching algorithm, which requires similarities of the iris data in the database with the captured one. In addition, nowadays using non-cosmetic or prescribed contact lenses becomes more popular and more preferred choice of many people, which makes the number of contact lens wearers significantly increases. These eyes with contact lenses add more complexity to the iris recognition process, since it can disturb the matching process which then affect the performance of the system. This situation has motivated this study to propose an iris recognition system that works for eyes with lenses. The proposed iris recognition system for eyes wearing non-cosmetic contact lenses was developed using circular hough transform (CHT) in the preprocessing phase for iris localization and discrete wavelet transform (DWT) for feature extraction. Experiments conducted on the proposed system has shown promising results with good accuracy of 0.95 for eyes with no contact lens and 0.8 for eyes with non-cosmetic lenses. The findings also suggested the importance of the iris localization process to the performance of the recognition system.

DIGITAL IDENTIFICATION AND FINANCIAL MONITORING: NEW TECHNOLOGIES IN THE FIGHT AGAINST CRIME

The purpose of the research is to explore the role of emerging technologies in combating criminal activities, particularly in digital identification and financial monitoring. The article aims to highlight how these new technologies can enhance the effectiveness of law enforcement agencies and financial institutions in detecting and preventing crime. Methods. The methodological basis of the research was a set of methods of scientific knowledge. In particular, the phenomenological method allowed the author to analyse national, international, and foreign law legal sources. The analytical method was used during the analysis of the conceptual and categorical apparatus, the definitions of “artificial intelligence”, “digital identification”, “blockchain”, etc. The author also used the method of analysis and generalisation, which made it possible to clarify and generalise views on the essence and types of digital identification for further use in financial monitoring. Results. The article discusses various technological advancements, such as biometric identification systems, blockchain technology, artificial intelligence, and data analytics, and their application in the fight against crime. It explores how digital identification systems can help verify individuals’ identities more securely and efficiently, reducing the risk of identity theft and fraud. Additionally, it examines how financial monitoring tools can enable authorities to track suspicious transactions, detect money laundering activities, and disrupt the financing of criminal organisations. The article also addresses potential concerns and challenges associated with using these technologies, such as privacy issues, data security, and ethical considerations. Conclusions. New technologies such as biometrics (e.g., fingerprints, facial recognition, iris scans) and blockchain-based systems can improve digital identification processes. These technologies offer increased security, reduced fraud, and streamlined identity verification procedures, preventing theft and other related crimes. Advanced data analytics, machine learning, and artificial intelligence algorithms can be utilised to enhance financial monitoring systems. These technologies enable faster and more accurate identification of suspicious transactions, money laundering activities, and financial crimes. Automated systems can analyse vast amounts of data and detect patterns that would be difficult for humans to identify, thereby assisting law enforcement agencies in combatting financial crimes.

Enhancing Voting Security and Efficiency

This study explores developing and implementing a novel Electronic Voting Machine (EVM) system integrated with biometric identifiers to enhance voting security and efficiency significantly. Traditionally, voting processes relied on paper ballots, a system fraught with several challenges, including over-voting, the loss or misplacement of ballot papers, environmental harm due to paper consumption, and a lengthy result compilation process. An advanced EVM system is proposed to address these issues, leveraging unique biometric identifiers - facial recognition and fingerprints - for voter authentication and secure vote recording. Our EVM system effectively improves the security against bogus voting and vote repetition, which have been significant concerns in previous voting systems. This robust approach to voter authentication minimizes the likelihood of voting fraud, thus contributing to a more reliable and secure voting process. However, the transition to this advanced EVM system is challenging. The study identifies keyimplications, including the impact on employment due to automation, potential inaccuracies and biases associated with biometric technologies, and vital privacy concerns surrounding using sensitive biometric data. Despite these challenges, the proposed system provides a substantial foundation for future enhancements. Opportunities for further development include the integration of additional biometric identifiers like iris recognition, refining the accuracy of current biometric technologies, and strengthening data privacy measures.