Our biometric technology in scientific papers
In addition to commercial software development, our fingerprint and face identification technologies are being used in a number of scientific studies and experiments in the biometrics and other fields. Below is a selection of research papers that present interesting ideas and methods for improving current biometric technologies.
J. J. Engelsma,
"Infant-ID: Fingerprints for Global Good".
arXiv.org preprint, 2010.03624 (2020).
Infant-Prints, an end-to-end infants' fingerprint recognition system, is proposed in the paper. System's accuracy was tested in a longitudinal scenario, where infants fingerprints were captured several times over a 12-month time span and matched between each other. VeriFinger template extraction and matching algorithm was used in a pipeline with the original software for testing the system.
"Towards More Accurate Contactless Fingerprint Minutiae Extraction and Pose-Invariant Matching".
IEEE Transactions on Information Forensics and Security, Vol. 15 (2020): pp. 3924-3937.
An original fingerprint minutiae extraction algorithm, intended for using with images obtained by contactless means, is proposed in the paper. VeriFinger 10.0 SDK template extraction algorithm was used in the accuracy tests together with the proposed algorithm. Also, the corresponding version of VeriFinger template matching algorithm was used in matching experiments with templates, which were generated by the proposed algorithm.
"Latent Fingerprint Image Enhancement Based on Progressive GenerativeAdversarial Network".
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) (2020).
A method of latent fingerprint enhancement, based on generative adversarial network (GAN), is proposed in the paper. The method makes focus on the local features such as minutiae, as well as preserves structure feature such as the orientation field and uses progressive training scheme. A subset of NIST SD14 dataset was used for training, and NIST SD27 dataset was used for testing and comparing the effectiveness of the method with other methods. VeriFinger algorithm was used for fingerprint feature extraction from the processed dataset images into biometric templates and matching the templates with the corresponding fingerprints from the datasets.
M. A. Acquah,
"Securing Fingerprint Template Using Blockchain and Distributed Storage System".
Symmetry, 12, 951 (2020).
A distributed, Blockchain-based system for storing, encryption and hashing of fingerprint templates is described in the paper. The authors claim that storing hashes of fingerprint templates in Blockchain provides secure method of ensuring that the templates are not modified by malicious actors. VeriFinger template extraction and matching algorithm was used as a part of the system.
C. Gottschlich, T. Hotz, R. Lorenz, S. Bernhardt, M. Hantschel, A. Munk.
Modeling the Growth of Fingerprints Improves Matching for Adolescents.
In IEEE Transactions on Information Forensics and Security, 2011.
A simple method of rescaling fingerprint images for predicting how would appear juvenile's fingerprints years later is presented in the paper. The proposed rescaling method was applied to a 3.25 million fingerprints database; the all-against-all verification test showed that the equal error rate (ERR) was ten times smaller compared to the original database. VeriFinger algorithm was used for template extraction and verification during the tests.
M.O.Derawi, B.Yang, C.Busch.
Fingerprint Recognition with Embedded Cameras on Mobile Phones.
A biometric method for mobile device user authentication is proposed in the paper. The technology tests were performed using two cellphones with cameras for making photos of fingerprints. The fingerprint database for the tests included all 10 fingerprints taken from 22 subjects. VeriFinger algorithm was used for template extraction and verification from the fingerprint images obtained with the cellphones.
C. Ryu, S.G. Kong, H.Kim.
Enhancement of Feature Extraction for Low-quality Fingerprint Images Using Stochastic Resonance.
The paper presents the usage of stochastic resonance for enchancing feature extraction from low-quality fingerprints. The results of experiments with fingerprint images from FVC2004 DB2 database showed that adding Gaussian noise to low-contrast fingerprint images allowed to extract features from most of these images. VeriFinger algorithm was used for fingerprint template extraction and verification.
C.Whitelam, Z.Jafri, T.Bourlai.
Multispectral Eye Detection: A Preliminary Study.
A problem of eye detection on a face across visible, multispectral and short-wave infrared band is studied in the paper. The experiments were performed using WVU Multispectral database and a subset of FRGC database, and showed promising results. VeriLook algorithm was used for comparison with the authors proposed method for eye detection on a face; the comparison showed similar reliability level.
H. Xu, R.N.J. Veldhuis.
Complex Spectral Minutiae Representation For Fingerprint Recognition.
In Computer Vision and Pattern Recognition Workshops (CVPRW), 2010 IEEE Computer Society Conference on.
The paper introduces complex spectral minutiae representation (SMC). SMC usage showed 2-4 times recognition accuracy improvement compared with location-based and orientation-based minutiae representations (SML and SMO). VeriFinger algorithm was used for fingerprint template extraction.
E. Marasco, A. Ross, C. Sansone.
Predicting Identification Errors in a Multibiometric System Based on Ranks and Scores.
In Proc. of 4th IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS), September 2010, Washington DC, USA. 1 - 6.
The paper presents a methology of improving the identification accuracy of multimodal biometric systems by using ranks and scores. A classifier for each modality has been designed using the ranks and scores in order to detect identification errors and produce rank-level fusion schemes in a multimodal scenario. VeriFinger and VeriLook algorithms were used for matching fingerprints and faces correspondingly.
J. Feng, A.K. Jain, A. Ross.
Detecting Altered Fingerprints.
In 20th International Conference on Pattern Recognition (ICPR 2010), 23-26 August 2010, Istanbul, Turkey. 1622-1625.
An algorithm for detecting altered fingerprints is proposed and tested in the paper. Experiments were conducted with real-world altered fingerprints and synthetic ones using the proposed algorithm and NFIQ software for comparison. VeriFinger engine was used for extracting skeletonized fingerprint images and orientation fields during the experiments.
R. Wang, X. Yang, X. Liu, S. Zhou, P. Li, K. Cao, J. Tian.
A Novel Fingerprint Template Protection Scheme Based on Distance Projection Coding.
In 20th International Conference on Pattern Recognition (ICPR 2010), 23-26 August 2010, Istanbul, Turkey. 886-889.
A fingerprint biometric cryptosystem scheme based on a codebook named distance projection for biometric coding used to generate secured biometric template is presented in the paper. The experiments on the FVC2002 database showed that the proposed scheme can obtain positive results on both security and authentication accuracy. VeriFinger algorithm was used for fingerprint minutiae extraction.
B. Yang, C. Busch, D. Gafurov, P. Bours.
Renewable Minutiae Templates with Tunable Size and Security.
In 20th International Conference on Pattern Recognition (ICPR 2010), 23-26 August 2010, Istanbul, Turkey. 878-881.
The paper proposes a renewable fingerprint minutiae template generation scheme to utilize random projection for template diversification in a security enhanced way. The scheme was tested with a database from FVC2002 and showed the desirable biometric performance. VeriFinger algorithm was used for fingerprint template extraction during the scheme development and tests.
Q. Zhao, J. Feng, A. K. Jain.
Latent Fingerprint Matching: Utility of Level 3 Features.
In MSU Technical Report. MSU-CSE-10-14, August 2010.
The paper studies the utility of level 3 features (pores, dots, incipient ridges, etc.) for latent fingerprint matching. Automatic algorithms for extracting and matching level 3 features are proposed. VeriFinger algorithm was used for fingerprint feature extraction and preliminary template matching using level 2 features (minutiae).
H. Xu, R.N.J. Veldhuis.
Binary Representations of Fingerprint Spectral Minutiae Features.
In 20th International Conference on Pattern Recognition (ICPR 2010), 23-26 August 2010, Istanbul, Turkey. 1212-1216.
The spectral minutiae representation is a method to represent a minutiae set as a fixed-length real-valued feature vector. Two novel methods of the spectral minutiae features quantization into binary strings (Spectral Bits and Phase Bits) were introduced to apply the spectral minutiae representation with a template protection scheme. The experiments on the FVC2002 database showed that the binary representations can even outperformed the spectral minutiae real-valued features. VeriFinger algorithm was used for fingerprint template extraction.
S. Yoon, J. Feng, A. K. Jain.
On Latent Fingerprint Enhancement.
In Proc of SPIE, Biometric Technology for Human Identification VII. April 2010.
A latent fingerprint enhancement algorithm, which required manually marked region of interest (ROI) and singular points, is proposed in the article. VeriFinger algorithm was used to generate a skeletonized fingerprint image for a given latent; the skeleton allowed to produce a coarse orientation field for comparison with the one generated by the proposed algorithm. Experimental results on NIST SD27 latent fingerprint database indicated that the proposed enhancement algorithm used together with the VeriFinger algorithm significantly improved latent fingerprint feature extraction accuracy.
Degtyarev N., Seredin O.
Comparative Testing of Face Detection Algorithms.
In A. E lmoataz et al. (Eds.): ICISP 2010, LNCS 6134, pp. 200--209, Springer, Heidelberg (2010).
Available at: lda.tsu.tula.ru/papers/degtyarev-2010-icisp-ctfd.pdf.
The original publication is available at www.springerlink.com.The study describes the testing of seven face detection algorithms; the results of their comparison are presented. The methologies of face detection algorithms quality evaluation and objective comparison was proposed. Each face detection algorithm was routinely tested in the labeled image datasets. VeriLook algorithm was among the tested and showed the best performance and highest face detection speed.
K. Manjunath, T. R. Saraswathi, G. Sriram, B. Sivapathasundharam, S. Porchelvam.
Reliability of Automated Biometrics in the Analysis of Enamel Rod End Patterns.
In J Forensic Dent Sci 2009;1(1):32-36.
The study researched the reliability and sensitivity of an automated biometric recognition software in analyzing tooth prints. During research, enamel rod end patterns were obtained three times from a specific area on the labial surface of ten extracted teeth using acetate peel technique. The acetate peels were subjected to analysis with VeriFinger SDK software to obtain the enamel rod end patterns (tooth prints) and respective minutiae scores for each tooth print. It was found that VeriFinger software was able to identify duplicate records of the same area of a same tooth with the original records stored on the database of the software. Comparison of the minutiae scores using Cronbach's test also showed that there was no significant difference in the minutiae scores obtained.
K. Manjunath, G. Sriram, T. R. Saraswathi, B. Sivapathasundharam.
Enamel Rod End Patterns: a Preliminary Study Using Acetate Peel Technique and Automated Biometrics.
In J Forensic Odontol 2008; 1(1):33-36.
In this study VeriFinger SDK software has been used to obtain tooth prints' (enamel rod end) pattern. A microphotograph of acetate peel imprint was used for biometric analysis. The study revealed that none of the different tooth prints showed intra- and inter-individual similarity.
S. Modi, S. Elliott, H. Kim.
Performance Analysis for Multi Sensor Fingerprint Recognition System.
In Lecture Notes in Computer Science, 2007, Volume 4812/2007, 292-305.
The article is available at SpringerLink.This research analyzes effects of interoperability on error rates for fingerprint datasets captured from two optical sensors and a capacitive sensor. A practical approach for improving performance of interoperable fingerprint datasets is presented. Several issues which need to be investigated to reduce the effects of interoperability on performance are revealed. VeriFinger fingerprint matching algorithm was used together with all three fingerprint scanners in the experiments.
Y. Chen, A. K. Jain.
Dots And Incipients: Extended Features for Partial Fingerprint Matching.
The paper proposes an algorithm to extract two major Level 3 feature types, dots and incipients, based on local phase symmetry and demonstrates their effectiveness in partial print matching. Since dots and incipients can be easily encoded by forensic examiners, the results of this research will probably have benefits to Next Generation Identification (NGI) systems. The VeriFinger matching algorithm was used in the experiments described in the paper.
A. Ross, R. Nadgir.
A Calibration Model For Fingerprint Sensor Interoperability.
In Proc. of SPIE Conference on Biometric Technology for Human Identification III, (Orlando, USA), pp.62020B-1 – 62020B-12, April 2006.
The paper proposes a simple non-linear calibration scheme, based on Thin Plate Splines (TPS), as sufficient to facilitate sensor interoperability in the context of fingerprints. The VeriFinger algorithm and NIST BOZORTH3 algorithm were used in the fingerprint matching experiments on improving identification results with images from two different fingerprint sensors.
T. Uz, G. Bebis, et.al.
Minutiae-Based Template Synthesis and Matching Using Hierarchical Delaunay Triangulations.
A new minutiae template-merging method is presented in the paper. The method is based on hierarchical Delaunay triangulations for synthesizing a super-template from multiple enrollment templates to increase coverage area, restore missing features, and alleviate spurious minutiae. The VeriFinger algorithm was used in the experiments for extracting the minutiae from fingerprints.
M. Arnold, C. Busch, H. Ihmor.
Investigating Performance and Impacts on Fingerprint Recognition Systems.
In Sixth Annual IEEE Systems, Man and Cybernetics (SMC) Information Assurance Workshop 2005. Proceedings: June 15 - 17, 2005, West Point, New York. Workshop papers. Piscataway, NJ: IEEE, 2005, pp. 1-7.
The article is available at IEEE Xplore.The paper presents a comparative study of fingerprint recognition systems with special emphasis on investigating the performance with regard to different combinations of fingerprint sensors and algorithms and the impact of ageing. The goal of this study was to investigate the capability characteristics of biometric systems regarding integration of biometric features in personnel documents such as IDcards and Visa application documents. Neurotechnology's fingerprint identification algorithm was used for testing among 6 other algorithms in conjunction with 11 different fingerprint scanners.
R. K. Rowe, K. A. Nixon, S. P. Corcoran.
Multispectral Fingerprint Biometrics.
In Sixth Annual IEEE Systems, Man and Cybernetics (SMC) Information Assurance Workshop, 2005. Proceedings: June 15 - 17, 2005, West Point, New York. Workshop papers. Piscataway, NJ: IEEE, 2005, pp. 14-20
The article is available at IEEE Xplore.A novel fingerprint sensor is described in the paper. The sensor combines a multispectral imager (MSI) with a conventional optical fingerprint sensor for improving usability and security relative to standard technology. VeriFinger algorithm was used in the experiments for fingerprint minutiae detection, extraction and matching.
T. C. Clancy, N. Kiyavash, D. J. Lin.
Secure Smartcardbased Fingerprint Authentication.
In Proceedings of the 2003 ACM SIGMM workshop on Biometrics methods and applications, November 08, 2003, Berkley, California.
A cryptosystem capable of using fingerprint data as the key is presented in the paper. The fundamental insecurities hampering a scalable, wide-spread deployment of biometric authentication are also examined with a focus on situations where a private key is stored on a smart card that is used for authentication in a networked environment. The VeriFinger algorithm was used to extract fingerprint features for the tests performed in this paper.
S. K. Modi, S. J. Elliott.
Impact of Image Quality on Performance: Comparison of Young and Elderly Fingerprints.
In Proceedings of the 6th International Conference on Recent Advances in Soft Computing (RASC 2006), K. Sirlantzis (Ed.), pp. 449-454, 2006.
This research examines the impact of fingerprint image quality of two different age groups: 18-25, and 62+ on overall performance using two different fingerprint matchers. The difference in image quality between the two age groups and the impact of image quality on performance of fingerprint matchers between the two groups were analyzed. Neurotechnology's VeriFinger and NIST bozorth3 matchers were used to assess overall performance.
R. F. Stewart, R. Youmaran, A. Adler.
Fingerprint Verification For Control Of Electronic Blast Initiation.
In Conf. of the Int. Society of Explosives Engineers., Nashville, TN, USA, January 2007.
The paper shows the advantages of employing fingerprint biometric technology to ensure that only authorized personnel can initiate a blast involving electronic detonators. MegaMatcher-based software and a fingerprint sensor were used in the application of electronic blast initiation control.
A. Mascher-Kampfer, H. Stögner, A. Uhl.
Comparison Of Compression Algorithms' Impact On Fingerprint And Face Recognition Accuracy.
In Visual Communications and Image Processing 2007 (VCIP'07), Proceedings of SPIE 6508, p. 650810-1 - 65080N-10.
The impact of using different lossy compression algorithms on the matching accuracy of fingerprint and face recognition systems is investigated in the paper. VeriFinger and VeriLook algorithms were used in the experiments for fingerprint and face recognition respectively.