Fingerprint Classification Research Articles

Vision transformer for contactless fingerprint classification

Vision transformer for contactless fingerprint classification

Universal Fingerprint Generation: Controllable Diffusion Model with Multimodal Conditions.

The utilization of synthetic data for fingerprint recognition has garnered increased attention due to its potential to alleviate privacy concerns surrounding sensitive biometric data. However, current methods for generating fingerprints have limitations in creating impressions of the same finger with useful intra-class variations. To tackle this challenge, we present GenPrint, a framework to produce fingerprint images of various types while maintaining identity and offering humanly understandable control over different appearance factors such as fingerprint class, acquisition type, sensor device, and quality level. Unlike previous fingerprint generation approaches, GenPrint is not confined to replicating style characteristics from the training dataset alone: it enables the generation of novel styles from unseen devices without requiring additional fine-tuning. To accomplish these objectives, we developed GenPrint using latent diffusion models with multimodal conditions (text and image) for consistent generation of style and identity. Our experiments leverage a variety of publicly available datasets for training and evaluation. Results demonstrate the benefits of GenPrint in terms of identity preservation, explainable control, and universality of generated images. Importantly, the GenPrint-generated images yield comparable or even superior accuracy to models trained solely on real data and further enhances performance when augmenting the diversity of existing real fingerprint datasets.

Enhanced fingerprint classification through modified PCA with SVD and invariant moments.

This research introduces a novel MOMENTS-SVD vector for fingerprint identification, combining invariant moments and SVD (Singular Value Decomposition), enhanced by a modified PCA (Principal Component Analysis). Our method extracts unique fingerprint features using SVD and invariant moments, followed by classification with Euclidean distance and neural networks. The MOMENTS-SVD vector reduces computational complexity by outperforming current models. Using the Equal Error Rate (EER) and ROC curve, a comparative study across databases (CASIA V5, FVC 2002, 2004, 2006) assesses our method against ResNet, VGG19, Neuro Fuzzy, DCT Features, and Invariant Moments, proving enhanced accuracy and robustness.

Development of a semi-supervised machine learning based noise filter for quantum cascade laser-coupled mid-infrared spectrometer

The rapid and accurate quantitative determination of molecular rovibrational spectral features is highly desirable for applications in fundamental high-resolution spectroscopy, trace chemical sensing, and environmental monitoring. But significant spectral noise can impede the accuracy and precision of such spectral measurements as well as the precise classification of fine and hyperfine spectral fingerprints. Nevertheless, the denoising of weak rovibrational spectral data has long been a computational and experimental challenge. Here, we develop a new semi-supervised machine learning (SSML) denoising method combining unsupervised spectral eigenvector space dimensionality reduction of the spectral data matrix exploiting principal component analysis with supervised Fourier domain residual analysis for enhancing high-resolution rovibrational spectral signal quality. We discussed the detailed implementation of the SSML algorithm on the real experimental mid-infrared rovibrational spectral features of nitrogen dioxide (NO2) in the gas-phase acquired from a quantum cascade laser-coupled cavity ring-down spectroscopic method. Our results confirm the robustness and feasibility of the SSML approach and could lead to a wide range of pertinent applications in infrared spectroscopy including precise spectral signal analysis, concentration retrieval and performance enhancement of laser-coupled spectroscopic sensors.

Klasifikasi Gender Berdasarkan Sidik Jari Menggunakan Principal Component Analysis dan Support Vector Machine

Most fingerprint classification research uses features such as core and delta as its basis. Before extracting fingerprint features, various preprocessing steps are usually performed first. This study differs from other studies in that classification is performed directly on the fingerprint image without going through a detailed preprocessing step and only a pixel size change is performed to 96x103. Fingerprint features are not determined manually, but are extracted automatically using the Principal Component Analysis (PCA) method which produces the 4200 best features. For feature perfection reasons, feature normalization has been performed using StandardScaler. The classification of this study uses a nonlinear Support Vector Machine (SVM) method with a Polynomial kernel. This study uses 6000 data samples from the SOCOFing database. This model obtains a classification accuracy of up to 88.75%.

Electronic anti-theft protection for vehicles of people with special needs

The article presents an electronic anti-theft protection subsystem that was incorporated into the vehicle control system dedicated to people with special communication needs. The aim of the work was to design and test an electronic subsystem allowing for the acquisition and quick analysis of fingerprint images, checking whether they belong to the vehicle owner and generating a signal allowing the vehicle to be started. The following hypothesis was put forward: it is possible to use artificial intelligence for accurate, automatic classification of fingerprint images acquired in a system with a single-chip microcomputer in order to identify a person. The research niche includes the use of biometrics in electronic anti-theft security adapted to people with special needs. This solution reduces the risk of theft and unauthorized use of special vehicles tailored to the owners' individual needs. The use of a commercial algorithm offered by the sensor manufacturer or the use of artificial neural networks for the classification of fingerprints was considered. The results of research on the accuracy of fingerprint recognition obtained from the developed subsystem are presented. The experiments performed based on the NASNetLarge artificial neural network model confirmed the possibility of achieving recognition accuracy for the test set of 99.99%. Additionally, practical aspects of the applicability of the presented device in an electric vehicle control system supporting the movement of people with physical and/or intellectual disabilities were discussed. The presented solution, after minor modifications, can be used in access security systems for protected rooms, as well as vehicles and military equipment.

ESTUDO DE IMPRESSÕES DIGITAIS LATENTES E NOVOS REVELADORES POR PAPER SPRAY MASS SPECTROMETRY (PS-MS)

STUDY OF LATENT FINGERPRINTS AND NEW DEVELOPERS BY PAPER SPRAY MASS SPECTROMETRY (PS-MS). In the criminal sphere, it is extremely important to identify perpetrators of crimes based on latent fingerprints (invisible to the naked eye) revealed at crime scenes and their corresponding objects. The present work aimed to evaluate four synthetic analogues of curcumin as possible IDL developers, and to compare their results with those of carbon black developer, as well as to reach the most adequate parameters in the use of the LTQ and FT-ICR mass spectrometers with the paper spray ionization (PS) source to study the chemical composition of the endogenous substances contained in the IDL. For this, a 24-1 factorial design was developed, where it was verified that the distance and voltage factors were the most influential for the increase in the response of signals, in the analysis of latent fingerprints by PS. Among the detected substances we can mention around 20 fatty acids, and squalene. Finally, four potential developers, synthetic analogues of curcumin, were compared with carbon black developer. The revelation of latent fingerprints occurred efficiently, allowing the visualization of minutiae, such as bifurcation and end of line, making it possible to classify them, according to the Vucetich Fingerprint Classification System.

Geomagnetic indoor localisation-based dilated convolution neural networks

ABSTRACT With the rapid development of mobile Internet, the demand for Indoor Location-Based Service (ILBS) keeps rising. Geomagnetic field fingerprinting-based positioning scheme, with low cost, high accuracy and good stability, has gradually earned the attention of researchers. There are many problems in the existing indoor localisation researches based on geomagnetic fingerprint recognition, such as heavy workload of fingerprint collection, insufficient positioning accuracy, high walking cost of positioning and long positioning delay. Therefore, a localisation system named DCGIL is proposed in this thesis. In DCGIL, a segmentation length estimation algorithm of geomagnetic fingerprint sequence is proposed to evaluate the minimum segmentation length, the shortest walking distance. Then, through an overlapping fingerprint segmentation method, the gap between coordinate points and the cost of fingerprint matching are reduced. Further, dilated convolutional neural networks is utilised for segmented fingerprint classification, which could improve the accuracy of fingerprint classification compared with the existing classification methods. Finally, DCGIL utilises the sliding window to align and match the segmented fingerprint. Experiments show that the localisation accuracy reached 1.21 metres at 3.8 m shortest walking distance derived from segmentation length estimation algorithm, which improved 28% compared to existing algorithms.

TGPrint: Attack fingerprint classification on encrypted network traffic based graph convolution attention networks

Nowadays, most network traffic is encrypted, which protects user privacy but hides attack traces, further hindering identifying attacks to inspect traffic packages. Machine Learning (ML) methods are widely applied to attack classification on encrypted traffic owing to no need for manual analysis. However, existing studies only concentrate on basic statistical features and cannot obtain the crucial attack behaviors hiding in the encrypted traffic. Worse still, attackers constantly update attack vectors to evade detection, which means outdated features extracted from historical traffic fail to recognize unseen attacks.As a solution, we propose an attack classification approach, attack fingerprint based on graphs of time-window (TGPrint). We first filter normal traffic flows using ML models to eliminate the impact of useless, noisy data for attack classification and maintain suspicious traffic. Then, we create attack graphs to depict interaction behaviors of attack-victim hosts from suspicious traffic containing crucial attack behaviors. Besides, we divide a specific duration for each attack to precisely elaborate attack graphs, where temporal, statistical, and aggregate features are extracted to portray attack behaviors. Finally, we utilize Graph Neural Networks (GNNs) to mine and grasp the crucial behavior patterns from attack graphs to generate fingerprints and classify attacks, even unseen attacks. Extensive experiments are conducted on well-known datasets to verify our approach. It achieves a precision of 99% in attack classification on encrypted traffic, an average higher than other ML methods of 50%. Meanwhile, it classifies unseen attacks with an average accuracy of over 80% and has a strong robustness to false positives.

An Investigation of the Relationship between Fingerprints and Anaerobic Powers of Sports Sciences Students

In this study, the relationship between fingerprints and the anaerobic power of athletes was analyzed in a random sample group in correlation type. Fingerprints have been used electronically for identification in forensic criminology and authentication in business and social life with the development of dermatoglyphics science in the last century and the understanding that fingerprints are unique to the individual. Today, it is known that much research is being carried out to determine genetic characteristics, heredity, gender, character, and ability analysis from fingerprints. In this study, the height and weight measurements of 126 athletes from Cumhuriyet University Faculty of Sports Sciences were taken with the appropriate sampling method, and the vertical jump test was applied to the individuals. The anaerobic power of the athletes was calculated with these collected data. The coding method determined 10 fingerprints of the same sample group, and fingerprint classes and attributes were determined by observation. According to the findings obtained from the data analysis, a significant difference was observed between the anaerobic powers of the athletes according to their fingerprint classes. The anaerobic power of athletes with W2 Normal fingerprint codes has been observed to be higher than those without W2. However, it has been observed that fingerprints in certain codes increase and decrease in direct proportion to anaerobic power. In light of the data obtained in this study, limited but meaningful data were obtained in the direction of detecting sportive skills from fingerprints.

Effects of solvent-based adhesive removal on the subsequent dual analysis of fingerprint and DNA

The combined approach of classical fingerprinting and DNA profiling is a powerful tool in forensic investigations of latent “touch” traces. However, little attention has been paid to the organic solvents frequently used in dactyloscopic laboratories to facilitate the separation of adhesive evidence prior to fingerprint development and downstream effects on subsequent DNA profiling. In the present study, we tested a selection of adhesive removers (n = 9) and assessed their potential impact on DNA recovery and amplification by PCR. Thereby, we identified and characterized novel PCR inhibitors. All investigated chemicals contain volatile organic compounds that evaporate under normal indoor atmospheric conditions. Exposure to certain solvents resulted in increased DNA degradation, but only if evaporation was prevented. A series of adhesive-removal experiments were conducted with prepared mock evidence (self-adhesive postage stamps affixed to paper envelope) to investigate the impact of treatment time and the location of applied traces on DNA recovery and dactyloscopy, respectively. Due to the early onset of print decomposition, we found that only a short treatment time was compatible with the development of fingerprints on the adhesive side of a stamp. Solvents also removed DNA from the adhesive surface, thus resulting in a marked shift in the substrate distribution of recovered DNA from the stamp to the envelope, but not in the reverse direction. Furthermore, we observed that treatment with conventional fingerprint reagents lead to a significant reduction in the amounts of DNA recovered from stamps, while the additional use of adhesive removers did not significantly enhance this effect.

Telegraphic code for fingerprints: How justice was denied to the innovator who helped ameliorate the criminal justice system

In the last decade of the nineteenth century, an Indian officer of the Bengal Police, sub-inspector Hem Chandra Bose (1867–1949) invented the telegraphic code system for fingerprints and published it in 1916. Sir Charles Stockley Collins of Scotland Yard, who is worldwide recognized as the originator of fingerprint telegraphic technique, published his findings in 1921—five years after Bose’s publication. Likewise Bose devised the single digit fingerprint classification system three years prior to Harry Battley, also of the Scotland Yard. The then British government honored Bose by awarding him an honorarium of Rs. 5000/- and by conferring the title of Rai Bahadur on him. However, this is not enough; Bose ought to be more widely recognized as a pioneer of the science of fingerprinting.

A multi-classifier system for automatic fingerprint classification using transfer learning and majority voting

A multi-classifier system for automatic fingerprint classification using transfer learning and majority voting

Integration of EfficientNetB0 and Machine Learning for Fingerprint Classification

Integration of EfficientNetB0 and Machine Learning for Fingerprint Classification

Classification of Manifold Learning Based Flight Fingerprints of UAVs in Air Traffic

As the number of UAVs (Unmanned Aerial Vehicles) and the market size have been expanding rapidly in recent years, projects such as NextGen and SESAR aim to include UAVs in air traffic. Therefore, different perspectives on understanding flight patterns can contribute to more effective management of future air traffic. Analysis of flight data offers an important insight into the operations of a UAV. In this study, it is aimed to extract a flight fingerprint using different machine learning techniques by means of a public dataset and the data obtained from our experimental flights. To get the individual flight pattern, multidimensional UAV sensor data has been reduced using manifold learning methods. By comparison, the most proper manifold method that allows highest classification accuracy (CA) has been investigated. Their performances are compared using both different manifold types and different classification methods. Then, the obtained manifold is used as flight fingerprints and validated by classification techniques. Various unsupervised manifold learning techniques such as t-Distributed Stochastic Neighbor Embedding (t-SNE), Locally Linear Embedding (LLE), Isometric Feature Mapping (ISOMAP) were tried for dimension reduction. For flight fingerprint classification, supervised machine learning techniques such as k-Nearest Neighbors (k-NN), Adaboost, Neural Network, Bayes, etc., were tested. It has been observed that the highest classification accuracy is achieved with the t-SNE manifold and k-NN classification pair. The extracted fingerprint can find many application areas such as performance tests in production lines, air traffic control, risk analysis, anomaly detection, observing pilot performance, drone efficiency over time.

A-Survey: Identification and Classification of Fingerprints via the Extreme Learning Machine Algorithm

The fingerprint comes to be the most popular and utilized biometric for identifying persons owing to its bio-invariant characteristic, precision, as well as easy acquisition. A sub-system of an identification system is the classification stage in order to diminish the penetration rate and computational complexity. Actually, there are many formal investigations regarding techniques by exploiting convolutional neural networks (CNN) together with fingerprint images, which have superior performance metrics at the cost of large training times even employing high-performance computing, which is not feasible in the standard world. In our manuscript, researches about identifying and classifying fingerprint databases by recurring to extreme learning machines (ELM) will be extensively reported and discussed for the first time. The diverse methodologies (ELM plus feature extractors) given by the authors will be studied and contrasted considering performance analysis. Consequently, academic papers with diverse versions of ELMs are developed to observe the pros and cons that they exhibit with each other and to probe how they may help for minimizing the penetration rate of fingerprint databases. In fact, this issue is very relevant because enhancing the penetration rate means shorting search times and computational complexity in fingerprints.

The History of Forensic Science in India by Saumitra Basu

Reviewed by: The History of Forensic Science in India by Saumitra Basu Saurav Kumar Rai (bio) The History of Forensic Science in India By Saumitra Basu. London and New York: Routledge, 2021. Pp. xvii + 204. "Forensic science," the application of science for purposes of law and justice, entails the uses of biology, physics, chemistry, and general scientific principles and methods to solve legal problems. Thus, legal setup and requirements play a crucial role in the evolution of forensic science. In India, argues Saumitra Basu in The History of Forensic Science in India, the rise, growth, and development of forensic science can be located in the colonial period (p. 19). The British Raj not only drastically altered the precolonial concept of crime and criminal justice but also revolutionized the nature and magnitude of criminal investigation. Extensive use of forensic science in criminal jurisprudence was part and parcel of this phenomenon. However, instead of beginning his narrative directly in the colonial period, the author starts with a discussion of the beginning of crime identification in different civilizations. Thus, in the first chapter, he begins with the notion of crime and punishment and associated investigations in ancient India with special attention focused on toxicology (pp. 27–56). Here, Basu's primary focus is on agada tantra, the branch of Ayurveda (a traditional Indian healing system) that deals with various kinds of toxins, ways to clean them out of the body, and antidotes for particular poisons. Similarly, when discussing crime and justice in ancient India, the prime attention is on the Arthashastra, a comprehensive administrative manual. However interesting these discussions are, they do not connect well with the central theme of the book, as forensic science was not used in ancient India (p. 45). Basu does, however, believe that crime was detected according to scientific observation and analysis, as delineated in agada tantra and the Arthashastra. This discussion seems to reveal the author's love for India's ancient past more than historical rigor. This assessment also applies to the second chapter, which looks at long-term developments in criminal investigation in West Asia, China, Rome, Egypt, England, America, and Latin America. In fact, investigation of any crime by its very nature includes the application of some scientific methods to reach any conclusion. However, it should not be treated as a precursor of forensic science, which is very much a modern discipline, even though the word "forensic" can be traced to the Latin word forum, or the public square where cases were heard and debated based on certain evidence. This is like [End Page 584] arguing for the premodern presence of biotechnology by taking examples of past practices of the fermentation of grains to produce beer. From chapter 3 onward, the author deals with the development of forensic science in colonial India. The colonial rulers were very proud of the introduction of the "rule of law" in India. The evolution of criminal justice in colonial India as a form of "governmentality" is discussed at length in the book (pp. 87–107). These legal reforms facilitated the growth of forensic science in India. Starting from the mid-nineteenth century, institutions were established in different parts of India to ensure scientific certainty in criminal investigations and subsequent trials at courts. It began with the use of mug shots, police photographs of detained persons. Then came the Chemical Examiners' Laboratories, Anthropometric Bureau, Finger Print Bureau, Explosive Department, Handwriting Department, and several others (ch. 4). Chapter 5 discusses fingerprinting as a unique case study of forensic practice. William Herschel, a British civil servant in India, was the first to advocate the use of fingerprinting in the identification of criminal suspects. Very soon this technique received official recognition. This led to the establishment of the world's first-ever fingerprint agency, in 1897 in Calcutta. Here Edward Richard Henry, the inspector-general of police, Lower Bengal, with the able assistance of two Indian sub-inspectors—namely, Azizul Haque and Hem Chandra Bose—developed a system of classification of fingerprints that is well documented by the author. This book is a novel attempt to explore the colonial origin of forensic science in India along with the role played by nonmedical...

Real-time bot infection detection system using DNS fingerprinting and machine-learning

In today’s cyberattacks, botnets are used as an advanced technique to generate sophisticated and coordinated attacks. Infected systems connect to a command and control (C&C) server to receive commands and attack. Thus, detecting infected hosts makes it possible to protect the network’s resources and prevent them from illicit activities toward third parties. This research elaborates on the design, implementation, and results of a bot infection detection system based on Domain Name System (DNS) traffic events for a network corporation. An infection detection feasibility analysis is performed by creating fingerprints. The traces are generated from a numerical analysis of 13 attributes. These attributes are obtained from the DNS logs of a DNS server. It looks for fingerprint anomalies using Isolation Forest to label a host as infected or not. In addition, on the traces cataloged as anomalous, a search will be carried out for queries to domains generated by Domain Generation Algorithms (DGA). Then, Random Forest generates a model that detects future bot infections on hosts. The devised system integrates the ELK stack and Python. This integration facilitates the management, transformation, and storage of events, generation of fingerprints, machine learning application, and analysis of fingerprint classification results with a precision greater than 99%.

Artificial Neural Network-Based Fingerprint Classification and Recognition

Artificial Neural Network-Based Fingerprint Classification and Recognition

Accelerating deep learning inference via layer truncation and transfer learning for fingerprint classification

AbstractBiometric systems have been used extensively in the identification and verification of persons. Fingerprint biometrics stands out as the most effective due to their characteristics of Permanence, uniqueness, ergonomics, throughput, low cost, and lifelong usability. By reducing the number of comparisons, biometric recognition systems can effectively deal with large‐scale databases. Fingerprint classification is an important task used to reduce the number of comparisons by dividing fingerprints into classes. Deep learning models have demonstrated impressive classification performance in fingerprint classification tasks. The high‐level features of deep learning models can affect the transfer learning in deep learning models. Furthermore, the high‐level features involve high computational costs that can render difficulty in the deployment of the applications. This work proposes an improved system for fingerprint classification through the truncation of layers and transfer learning. Our approach modifies the ResNet50 model to improve its network inference speed and performance in fingerprint classification by removing some deep convolutional layers. We then finetune the modified model and train it using a fingerprint dataset. The results show that the finetuned modified model improves classification accuracy at a reduced computational cost. At only 5.1M parameters, our model obtained a classification accuracy of 93.3% and precision of 93.4% performing better than previous studies based on its size‐performance ratio.