Fingerprint Identification System Research Articles

Multi-Layered Minutiae Extraction Based on Fusion-Attention for OCT Fingerprints

Minutiae extraction plays an important role in automated fingerprint identification systems. There are many minutiae extraction methods proposed for various fingerprints (rolled/slap/latent) in the past decades. However, for fingerprints by optical coherence tomography (OCT) imaging which have the essential property of multi-layered tissue structure, previous work may not be suitable. This paper is devoted to extract minutiae for OCT-based fingerprints based on the characteristics of OCT imaging and powerful deep neural networks. Specifically, we propose a multi-tissue-layered minutiae extraction network based on fusion-attention, involving a multi-tissue-layered feature extractor to fully explore the rich information in multiple subsurface fingerprints, and a one-stage detection framework to detect minutiae points. To demonstrate the effectiveness and robustness of the proposed method, a series of experiments have been carried out on the public OCT fingerprint benchmark dataset (https://github.com/FengLiu-0013/). Our method obtains the best results with an improvement of 8.3% in F1-score compared with state-of-the-art methods. The proposed method has also been evaluated for fingerprint identification using the extracted minutiae. Our method has achieved a matching error rate of 1.57%, with a relative reduction of 41% compared with state-of-the-art methods.

An Innovative Possibilistic Fingerprint Quality Assessment (PFQA) Filter to Improve the Recognition Rate of a Level-2 AFIS.

In this paper, we propose an innovative approach to improve the performance of an Automatic Fingerprint Identification System (AFIS). The method is based on the design of a Possibilistic Fingerprint Quality Assessment (PFQA) filter where ground truths of fingerprint images of effective and ineffective quality are built by learning. The first approach, QS_I, is based on the AFIS decision for the image without considering its paired image to decide its effectiveness or ineffectiveness. The second approach, QS_PI, is based on the AFIS decision when considering the pair (effective image, ineffective image). The two ground truths (effective/ineffective) are used to design the PFQA filter. PFQA discards the images for which the AFIS does not generate a correct decision. The proposed intervention does not affect how the AFIS works but ensures a selection of the input images, recognizing the most suitable ones to reach the AFIS's highest recognition rate (RR). The performance of PFQA is evaluated on two experimental databases using two conventional AFIS, and a comparison is made with four current fingerprint image quality assessment (IQA) methods. The results show that an AFIS using PFQA can improve its RR by roughly 10% over an AFIS not using an IQA method. However, compared to other fingerprint IQA methods using the same AFIS, the RR improvement is more modest, in a 5-6% range.

Statistics of fingerprint minutiae frequency and distribution based on automatic minutiae detection method

The Daubert case in Philadelphia in 1999 caused a debate about the scientificity of fingerprint evidence. Since then, the current fingerprint identification system has been constantly challenged and questioned. Quantitative identification technology based on the statistics of fingerprint minutiae has become a new research hot spot. In this paper, an automatic detection algorithm is designed to achieve automatic classification of fingerprint minutiae using the deep convolution neural network YOLOv5 model. Then the occurrence frequencies of minutiae are statistically evaluated in 619,297 fingerprint images. The results show that the frequency ranges (unit%) of six types of minutiae per finger are ridge endings [68.49, 70.81], bifurcations [26.37, 27.26], independent ridges [1.533, 1.626], spurs [1.129, 1.198], lakes [0.4588, 0.4963], crossovers [0.3034, 0.3256]. The results also show that there are differences in the distribution frequency of the six types of minutiae in the ten finger positions ( thumb, middle, ring, index and little finger of the left and right hand) and in the four finger patterns ( arch, left loop, right loop and whorl). From the quantitative point of view of fingerprint identification, this paper calculates the number and frequency ranges of six types of minutiae, distinguishes the evaluation value of each type of minutiae, and provides the basic data support for establishing a probability model of fingerprint identification in the future.

Research on the local regional similarity of automatic fingerprint identification system fingerprints based on close non‐matches in a ten million people database—Taking the central region of whorl as an example

The local regional similarity of fingerprints has always been a hot issue in the field of fingerprint research. With the increasing size of ten-print databases, the appearance of close non-matches (CNMs) in automatic fingerprint identification system (AFIS) candidate lists has attracted increasing attention from forensic science departments worldwide. In this study, three categories (high-, medium-, and low-level) of standards for CNMs were established and 60 whorl samples were marked with different numbers of minutiae to explore the occurrence and influencing factors of CNMs in AFIS candidate lists based on a ten million people database. The results showed that all prints could be found with their corresponding CNMs. The average occurrence rate of CNMs for every query was 52.7% in the top 100 lists, and the most similar CNM was exactly the same in the local area of 12 coincidence points. CNMs appeared more in the middle and lower parts of the central region of the whorl. Moreover, shorter C2C distances and the same finger number and hand led to more CNMs being inspected. CNMs with higher similarity required a more extensive regional area and smaller minutiae density. We concluded that CNMs have a high occurrence rate in large-scale databases and many factors are closely related to them. Fingerprint examiners and researchers need to strengthen their understanding of CNMs to avoid the occurrence of misidentification like the Madrid bombings.

Investigating the impact of thresholding and thinning methods on the performance of partial fingerprint identification systems: a review

A fingerprint identification system is an application of pattern recognition and image processing. The performance of a fingerprint-based biometric system relies on pre-processing techniques employed on fingerprint images. Especially, thresholding and thinning methods are used to detect minutiae points representing local features and are often utilized to identify a person uniquely. However, studies on partial fingerprints exposed the MasterPrint vulnerability for partial fingerprint identification systems wherein the system performs non-unique user identification. The thresholding and thinning techniques may lead to spurious minutiae generation and stimulate huge MasterPrints. Here, we investigate the impact of thresholding and thinning methods on identification accuracy and the percentage of MasterPrint generated using a partial fingerprint identification method. The experiments comprise four thresholding methods, namely, iterative optimal thresholding, Otsu’s global image thresholding, Niblack local thresholding, and Bernsen’s local image thresholding. Furthermore, it employs four thinning methods, namely, Khalid, Mariusz, Marek thinning algorithm, Khalid, Marek, Mariusz, Marcin thinning algorithm, Hilditch thinning algorithm, and Stentiford thinning algorithm. The results demonstrate that the identification accuracy and percentage of MasterPrint generated varies significantly by replacing the underlying pre-processing methods. Consequently, each combination of thresholding and thinning methods might not be suitable for user identification in high-security applications using a partial fingerprint identification method. The investigation outcomes provide the guidelines to demonstrate the robustness of a partial fingerprint identification method.

Fingerprint Recognition Using Deep Learning - A Review

There have been efforts to address the problems with fingerprint identification systems that require physical contact by creating contactless fingerprint identification systems. Numerous studies on various aspects of contactless fingerprint processing, including the use of deep learning in various algorithmic frameworks, classical image processing, and the machine-learning pipeline, have been published. It was demonstrated that the deep learning-based solutions were more accurate than the alternatives. This effort was driven by a desire to provide a thorough assessment of these successes and their identified limitations. This study examined three approaches to contactless fingerprint recognition: (i) methods for capturing images of the fingerprint, (ii) traditional preprocessing techniques for enhancing fingerprint images for recognition tasks, and (iii) deep learning. (i) taking a picture of your finger, and (ii) using conventional image processing to get the picture ready for recognition. In total, eight research papers were found to meet both the inclusion and exclusion criteria. Based on this review's findings, we discussed the potential benefits of deep learning methods for biometrics and the challenges that still need to be overcome before these methods can be used in practical biometric settings.

Prototype-Guided Autoencoder for OCT-Based Fingerprint Presentation Attack Detection

Anti-spoofing ability is vital for fingerprint identification systems. Conventional fingerprint scanning devices can only obtain information from the fingertip surfaces, and their performance is susceptible to skin conditions and presentation attacks (PAs). However, optical coherence tomography (OCT) can scan subcutaneous tissue and obtain 3D fingerprint structures, naturally enhancing its PA detection (PAD) ability from the perspective of hardware. Existing unsupervised PAD methods are based on image reconstruction. However, the reconstruction error is easily affected by OCT noise and the rich details of OCT images. Therefore we propose feature-based reconstruction to alleviate this problem, called the prototype-guided autoencoder. The model consists of a memory module and a denoising autoencoder without the requirement of PA fingerprints. As only bona fide fingerprints are available during the training phase, the memory module contains the prototype features of the bona fide fingerprints. During the inference phase, as the prototype memory module is frozen, the reconstructed representation of the bona fide input is close to the bona fide fingerprint features. Calculating the distance between the original features and the prototype reconstructed representation of the sample can achieve PAD. To obtain a better decision making boundary, we propose a representation consistency constraint, which reduces the bona fide representation reconstruction distance closer, so that it is easier to differentiate between fingerprints and PAs.

Synthesis of Multi-View 3D Fingerprints to Advance Contactless Fingerprint Identification.

Billions of contact-based fingerprint images have been acquired in large databases. Contactless 2D fingerprint identification systems have emerged to provide more hygienic and secured alternatives and are highly sought under the current pandemic. The success of such an alternative requires high match accuracy, not just for the contactless-to-contactless but also for the contactless-to-contact-based matching, which is currently below expectations for large-scale deployments. We introduce a new approach to advance such expectations on match accuracy and also to address privacy-related concerns, e.g. recent GDPR regulations, in the acquisition of very large databases. This paper introduces a novel approach for accurately synthesizing multi-view contactless 3D fingerprints to develop a very large-scale multi-view fingerprint database, and corresponding contact-based fingerprint database. A unique advantage of our approach is the simultaneous availability of much-needed ground truth labels and alleviation of laborious and often prone to erroneous tasks performed by human labeling. We also introduce a new framework that can not only accurately match contactless to contact-based images but also contactless to contactless images, as both of these capabilities are simultaneously required to advance contactless fingerprint technologies. Our rigorous experimental results presented in this paper, both for within- database and cross- database experiments, illustrate outperforming results to simultaneously meet both of these expectations and validate the effectiveness of the proposed approach.

Multifunctional Eu(III)-modified HOFs: roxarsone and aristolochic acid carcinogen monitoring and latent fingerprint identification based on artificial intelligence.

The exploration of multifunctional materials and intelligent technologies used for fluorescence sensing and latent fingerprint (LFP) identification is a research hotspot of material science. In this study, an emerging crystalline luminescent material, Eu3+-functionalized hydrogen-bonded organic framework (Eu@HOF-BTB, Eu@1), is fabricated successfully. Eu@1 can emit purple red fluorescence with a high photoluminescence quantum yield of 36.82%. Combined with artificial intelligence (AI) algorithms including support vector machine, principal component analysis, and hierarchical clustering analysis, Eu@1 as a sensor can concurrently distinguish two carcinogens, roxarsone and aristolochic acid, based on different mechanisms. The sensing process exhibits high selectivity, high efficiency, and excellent anti-interference. Meanwhile, Eu@1 is also an excellent eikonogen for LFP identification with high-resolution and high-contrast. Based on an automatic fingerprint identification system, the simultaneous differentiation of two fingerprint images is achieved. Moreover, a simulation experiment of criminal arrest is conducted. By virtue of the Alexnet-based fingerprint analysis platform of AI, unknown LFPs can be compared with a database to identify the criminal within one second with over 90% recognition accuracy. With AI technology, HOFs are applied for the first time in the LFP identification field, which provides a new material and solution for investigators to track criminal clues and handle cases efficiently.

Toward better AFIS practice and process in the forensic fingerprint environment

Automated Fingerprint Identification Systems (AFIS) are used in forensic operational environments worldwide. Traditionally, these systems are search systems only and have limited workflow to support forensic processes in recommended best-practice approaches. This paper addresses this issue by presenting best practice approaches and processes for AFIS in forensic environments. The discussion is divided into three parts. First, we identify three factors that can impact the performance of AFIS in forensic environments. Second, we discuss the search process and highlight how the National Police of the Netherlands (NPN) strategies to mitigate bias and error in their AFIS workflow. Finally, we briefly discuss other considerations for establishing best-practice. We offer resources for biometric managers and users, providing an overview of the best AFIS practices in forensic environments. We propose the NPN system as a benchmark and aim to foster discussions among AFIS communities to enhance AFIS practices and processes in forensic fingerprint environments.

Automated Fingerprint Identification System: with and without the Possibility of Correction of a Digitalised Image

Automated Fingerprint Identification System: with and without the Possibility of Correction of a Digitalised Image

Information loss, contextual information, and distinctiveness influence how well novice analysts discriminate fingerprints

AbstractAcross two experimental studies, we investigated the role of information loss, contextual information, and distinctive features of fingerprints on novice's ability to judge whether two fingerprints came from the same source. Distinctive fingerprints resulted in more accurate decisions. Information loss diminished performance on the comparison task, likely due to the loss of distinctive features. When people received contextual information from a fingerprint matching software (Automated Fingerprint Identification System or AFIS) suggesting that the two fingerprints were from the same source, they were less accurate and more biased to render a “same source” decision. These findings suggest that analysts should evaluate the characteristics of a fingerprint prior to engaging in a comparison task, and that fingerprints lacking information, particularly distinctive characteristics, should be more cautiously evaluated.

Optimized LightGBM Power Fingerprint Identification Based on Entropy Features.

The huge amount of power fingerprint data often has the problem of unbalanced categories and is difficult to upload by the limited data transmission rate for IoT communications. An optimized LightGBM power fingerprint extraction and identification method based on entropy features is proposed. First, the voltage and current signals were extracted on the basis of the time-domain features and V-I trajectory features, and a 56-dimensional original feature set containing six entropy features was constructed. Then, the Boruta algorithm with a light gradient boosting machine (LightGBM) as the base learner was used for feature selection of the original feature set, and a 23-dimensional optimal feature subset containing five entropy features was determined. Finally, the Optuna algorithm was used to optimize the hyperparameters of the LightGBM classifier. The classification performance of the power fingerprint identification model on imbalanced datasets was further improved by improving the loss function of the LightGBM model. The experimental results prove that the method can effectively reduce the computational complexity of feature extraction and reduce the amount of power fingerprint data transmission. It meets the recognition accuracy and efficiency requirements of a massive power fingerprint identification system.

Contactless Fingerprint Recognition Using Deep Learning—A Systematic Review

Contactless fingerprint identification systems have been introduced to address the deficiencies of contact-based fingerprint systems. A number of studies have been reported regarding contactless fingerprint processing, including classical image processing, the machine-learning pipeline, and a number of deep-learning-based algorithms. The deep-learning-based methods were reported to have higher accuracies than their counterparts. This study was thus motivated to present a systematic review of these successes and the reported limitations. Three methods were researched for this review: (i) the finger photo capture method and corresponding image sensors, (ii) the classical preprocessing method to prepare a finger image for a recognition task, and (iii) the deep-learning approach for contactless fingerprint recognition. Eight scientific articles were identified that matched all inclusion and exclusion criteria. Based on inferences from this review, we have discussed how deep learning methods could benefit the field of biometrics and the potential gaps that deep-learning approaches need to address for real-world biometric applications.

A Scattering Wavelet Network-based Approach to Fingerprint Classification

In a large-scale automatic fingerprint identification system (AFIS), fingerprint classification is an essential indexing step to reduce the search time in a large database for accurate matching. Fingerprint classification is still a challenging machine learning problem due to large intra-class and small inter-class variability. Nonlinear elastic deformation is one of the main sources of intra-class variability, which occurs due to the non-uniform pressure applied during fingerprint acquisition and the elastic nature of the fingerprint itself. This paper proposes a novel approach to fingerprint classification based on a scattering wavelet network to extract translation and small deformation invariant local features. The resulting sparse invariant feature vectors are used as input to a simple generative PCA affine classifier for the classification. The experiments evaluated with two different protocols on a benchmark NIST SD-4 database show the effectiveness and robustness of the proposed fingerprint classification algorithm in terms of classification accuracy.

FLD-SRC: Fingerprint Liveness Detection for AFIS Based on Spatial Ridges Continuity

Automatic fingerprint identification system (AFIS) uses fingerprint to authenticate users, which is legal if the user is enrolled. However, numerous studies reveal that it is susceptible to spoofing attacks where a third person might freely synthesize counterfeit fingerprints to trick the scanner. To resist spoofing attacks, it makes fingerprint liveness detection (FLD) highly desirable. Most of previous work was to directly input the whole fingerprints into convolutional neural network, making it impossible to fully explore the relationship of spatial ridges, especially those with the latent fine-grained minutia on fingerprint ridges. Accordingly, in this paper, we exploit the relationship of spatial ridges in fingerprints and propose a novel FLD method based on spatial ridges continuity (FLD-SRC). Several fingerprint patches are first selected utilizing ridge texture saturation, and then uniformly split into several slices and thus construct the spatial continuity between pixels and between slices. Next, the proposed FLD-SRC learns deep features from fingerprints and eliminates redundant information. After that, the extracted feature maps are treated as a sequence and analyzed the intra-continuity by cascade gated recurrent unit (GRU). A discriminant slice grouping subnetwork is then developed to model the correlation between ridges slices and implicitly discover the discriminant inter-continuity. Pruning strategy is further utilized to reduce network parameters and promote its practical application in real scenarios. Experimental results, evaluated on three publicly available datasets, show the competitiveness of our method. Furthermore, in addition to reducing computational complexity, our method also shows the best ACE performance in cross-material and cross-sensor cases.

Pattern Mathematical Model for Fingerprint Security Using Bifurcation Minutiae Extraction and Neural Network Feature Selection

Biometric based access control is becoming increasingly popular in the current era because of its simplicity and user-friendliness. This eliminates identity recognition manual work and enables automated processing. The fingerprint is one of the most important biometrics that can be easily captured in an uncontrolled environment without human cooperation. It is important to reduce the time consumption during the comparison process in automated fingerprint identification systems when dealing with a large database. Fingerprint classification enables this objective to be accomplished by splitting fingerprints into several categories, but it still poses some difficulties because of the wide intraclass variations and the limited interclass variations since most fingerprint datasets are not categories. In this paper, we propose a classification and matching fingerprint model, and the classification classifies fingerprints into three main categories (arch, loop, and whorl) based on a pattern mathematical model using GoogleNet, AlexNet, and ResNet Convolutional Neural Network (CNN) architecture and matching techniques based on bifurcation minutiae extraction. The proposed model was implemented and tested using MATLAB based on the FVC2004 dataset. The obtained result shows that the accuracy for classification is 100%, 75%, and 43.75% for GoogleNet, ResNet, and AlexNet, respectively. The time required to build a model is 262, 55, and 28 seconds for GoogleNet, ResNet, and AlexNet, respectively.

Amorphous calcium phosphate nanoparticles allow fingerprint detection via self-activated luminescence

Herein, stable activator-free photoluminescent nanoparticles based on amorphous calcium phosphate (ACP) were obtained for the first time, and their performance for latent fingerprint imaging was investigated. ACP nanoparticles with irregular rounded shape and diameters in 10–40 nm range were prepared by a rapid and simple chemical precipitation followed by heat treatment at 400 °C for 4 h. Notably, diffuse reflectance spectroscopy, photoluminescence excitation and emission measurements reveal a high density of localized energy states within the wide optical band gap of heat-treated ACP (Eg = 5.25–5.42 eV). This behavior allowed the excitation of ACP in near-ultraviolet region (λexc = 450 nm, 2.75 eV), leading to an intense defect-related broadband (490–890 nm) photoluminescence emission centered at 540 nm (2.30 eV), 50 times more intense than untreated nanoparticles. The effect of lattice shrinkage due to structural water elimination, the presence of ionic vacancies (VCa and VO in PO43-) and carbonate groups (CO32-), on the luminescent properties of ACP were discussed in detail, as well as investigated after aqueous-mediated ACP crystallization into hydroxyapatite (HA). ACP nanoparticles were non-cytotoxic, as determined by MTT assay in Human Dermal Fibroblast neonatal (HDFn) cell line, with cell viabilities superior to 95% in all tested concentrations (20–320 μg/mL), after incubation for 24 and 48 h. Latent fingerprint images were obtained using the ACP nanoparticles under near-ultraviolet irradiation (λexc = 450 nm) in tweezers and LCD smartphone and successfully validated by the Integrated Automated Fingerprint Identification System used by Scientific Police in Spain. The present results evidenced that the new luminescent ACP nanoparticles are safe to be used and agree with the forensic requirements for future legal actions.

Fingerprint Liveness Detection Schemes: A Review on Presentation Attack

ABSTRACT Extant literature has highlighted the vulnerability of Automatic Fingerprint Identification System (AFIS) to various forms of attacks, indicating presentation attack as the most predominant. This form of attack involves the malicious utilization of ubiquitous materials such as silicone, gelatin, playdoh, among many others, to fabricate synthetic fingerprints to circumvent AFIS. As a result, various studies have posited some countermeasures, encompassing the use of hardware- based and software-based techniques. The hardware-based methods necessitate the integration of supplementary sensors for capturing other live human traits such as pulse rate, odour etc. Conversely, the software-based methods are focused on feature extraction and deep learning schemes. However, despite the robustness of the software-based approach as compared to the hardware-based technique, both schemes are still faced with an immense challenge in developing fingerprint spoof generalized models to mitigate the concern of cross-material (novel) detection. As a result, various studies have highlighted that the issue of fingerprint spoofing should be treated an” opened-set problem” (training only live fingerprints), rather than a” closed-set problem” (training live & spoof), birthing the development of fingerprint one-class classifiers. This article presents a comprehensive detail on fingerprint spoofing, extant countermeasures and the challenge still faced by fingerprint spoof detectors.

Next Generation Identification system: Latent print matching algorithm and casework practices

The development of automated fingerprint identification systems (AFIS) revolutionized casework in the friction ridge community. Through the evolution of AFIS development, algorithms have become increasingly sensitive and specific. Internal validation studies conducted by the Federal Bureau of Investigation Laboratory Division’s Latent Print Units (LPU) assessed the performance of the Next Generation Identification (NGI) system’s latent print matching algorithm under the various encoding methods of the Multi-Biometric Identification System (MBIS) platform for the purpose of casework implementation. Based on the results of these studies, casework practices in the LPU have become more streamlined and efficient. The increase in accuracy of the latent print matching algorithm has reduced the number of candidates compared per search. For fingerprints, the results showed the best indicator for the correct mate being the number one candidate was the difference in score between the first and second candidate. When the score difference was 1200 or higher, the first candidate was mated in 99.3% of the trials. This understanding may allow for additional quality assurance measures to be implemented in casework.