Mobile Touchless Fingerprint Recognition: Implementation ...
Touchless and less-constrained 3D fingerprint recognition
-
Upload
angelo-genovese -
Category
Engineering
-
view
327 -
download
1
Transcript of Touchless and less-constrained 3D fingerprint recognition
Touchless and Less-Constrained3D Fingerprint Recognition
Angelo Genovese
Università degli Studi di MilanoDepartment of Computer Science
via Bramante 65, I-26013 Crema (CR), [email protected]
December 21, 2015
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Outline
• Biometricso Traditional biometric systemso Unconstrained and less-constrained biometrics
• Fingerprint biometrics• Touchless fingerprint biometrics
o Touchless fingerprint imageo Possible applications of touchless fingerprint biometricso State of the art of touchless 3D fingerprint methods
• Proposed touchless 3D fingerprint recognitiono Methodo Experiments
• Computation of synthetic 3D fingerprint samples• Touchless 3D reconstruction of ancient fingerprints• Conclusions
2
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Biometrics
• Traditional recognition methods:o Key, password, smartcard, token
• Biometrics:
o Behavioral Voice
Gait
Signature
Keystroke
o Physiological Fingerprint
Iris
Hand geometry
Palmprint
Palmevein
Ear
ECG
DNA
0 0.5 1 1.5 2 2.5 3 3.5 4-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Time (seconds)
x
Segnale
Immagine originale + minuzie NIST (solo per controllare se calcola
3
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Traditional biometric systems
• Low usability and user acceptance:o Complex and highly cooperative acquisition procedures
o Can be perceived as privacy invasive
4
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Unconstrained andless-constrained biometrics
• Unconstrained biometrics
o Uncooperative subjects
o Uncontrolled scenarios
• Less-constrained biometrics
aim at using samples captured
o Contactless
o Higher distances
o Natural light conditions
o On the move
o …
5
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Fingerprint biometrics
• The most used biometric trait:o High distinctivity
o High permanence
• Contact-based sensors:o Low usability and user acceptance
o Images with non-linear distortions and low contrast regions
o Latent fingerprint on the sensor platen
o Sensibility to dust and dirt
6
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless fingerprint images
TouchlessTouch-based
R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, "Touchless fingerprintbiometrics: a survey on 2D and 3D technologies", in Journal of InternetTechnology, May, 2014.
7
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Possible applications oftouchless fingerprint biometrics
8
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless 3D fingerprint:state of the art
• 3D reconstruction:o Mosaickingo Structured lighto Multiple viewso Absorbed light
• Unwrapping methods:o Parametric models (e.g. cylinder, sphere, set of rings)o Non-parametric models based on minimization functions
9
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless fingerprint:some existing systems (1/4)
Mosaicking
10
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless fingerprint:some existing systems (2/4)
Structured light
11
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless fingerprint:some existing systems (3/4)
Multiple views
12
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless fingerprint:some existing systems (4/4)
Absorbed light
13
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless 3D fingerprint recognition (1/2)
R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, "Toward unconstrainedfingerprint recognition: a fully-touchless 3-D system based on two views on themove", in IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2015.
14
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless 3D fingerprint recognition (2/2)
• Pros:o Less-constrained
o Compensate for rotations
o Absence of distortions in the fingerprint images due to different pressures of the finger on the sensor
o More robust to dust and dirt
o More user acceptance
o Possibility to use the recognition methods in mobile devices with standard CCD cameras
• Cons:o Longer computational time
o Interoperability to be further studied
15
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Contactless acquisition (1/4)
16
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Contactless acquisition (2/4)
Camera A Camera B
17
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Contactless acquisition (3/4)
18
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Contactless acquisition (4/4)
19
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Preprocessing
20
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Segmentation
21
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
3D reconstruction
• Extraction and matching of the reference points.
• Refinement of the pairs of corresponding points.
• 3-D surface computation and image wrapping.
22
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Extraction and matching of the reference points (1/2)
23
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Extraction and matching of the reference points (2/2)
24
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Refinement of the pairs of corresponding points
• Based on Thin Plate Spline
25
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
3D surface computationand image wrapping
1. Triangulation
2. Linear interpolation
26
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Computation oftouch-equivalent images
• EnhancementoBackground subtraction
oNon-linear equalization (logarithm)
oButterworth low-pass filter
• Two-dimensional mapping
27
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Two-dimensional mapping (1/2)
• Enrollment:oCompensate for rotations
oComputation of 𝑁𝑅 rotations
28
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Two-dimensional mapping (2/2)
29
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Template computation
• Neurotechnology VeriFingeroCommercially available
oDesigned for touch-based images
30
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Matching
• Database entryo𝑁𝑅 templates 𝑇𝑒oOne for each rotation
• Live sampleo1 template 𝑇𝑓
31
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Experimental results (1/2)
• Scenario evaluationo Touch-based and touchless acquisitions
32
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Experimental results (2/2)
• Datasets description
• Accuracy of 3D reconstruction
• Recognition performance
• Robustness to finger misplacements
• User acceptability
• Interoperability
• Overview of different technologies
33
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Datasets description
• Touchless - one sessiono 2368 sampleso 10 fingers, 30 volunteers, 8 acquisitions per finger
• Touchless - two sessionso 2368 sampleso 10 fingers, 15 volunteers, 16 acquisitions per finger
8 acquisitions one year, 8 acquisition subsequent year
• Touchless - misplaced fingerso 1200 sampleso 2 fingers (index), 30 volunteers, 20 acquisitions per finger
• Touch-basedo One sessiono Two sessions
34
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Accuracy of 3D reconstruction
• Average error: 0.03m
35
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Recognition performance (1/2)
36
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Recognition performance (2/2)
• Comparable to touch-based systemsoOne session
o Two-session
37
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Robustness to finger misplacements
• Genuine and impostor match scores remainwell separated
38
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
User acceptability
• Survey performed using questionnaires
• Results show preference towards contactless recognition
39
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Interoperability
• Accuracy level obtained by matching imagescaptured by different devicesoMatching touchless with touch-based images
2 803 712 identity comparisons
EER = 2.00% with 𝑁𝑅 = 25
Less than EERs obtained in the literature with similar experiments
40
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Overview of the different technologies
Aspect Touch-based Touchless
Accuracy EER = 0.03% EER = 0.06%
Scalability High To be further investigated
Interoperability High To be improved
Security Latent fingerprints No latent fingerprints
Privacy Data protection techniques Data protection techniques
Cost 10$ to 5000$ 0$ to 5000$
Usability Medium High
User acceptance Medium High
Speed Template extraction + matching
3D reconstruction + template extraction + matching
41
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Computation of synthetictouchless fingerprint samples (1/2)
42
R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, "Accurate 3D fingerprintvirtual environment for biometric technology evaluations and experimentdesign", Proc. of the 2013 IEEE International Conference on ComputationalIntelligence and Virtual Environments for Measurement Systems andApplications (CIVEMSA 2013), Milan, Italy, July 15-17, 2013, pp. 43-48.
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Computation of synthetictouchless fingerprint samples (2/2)
43
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Touchless 3D reconstructionof ancient fingerprints
44
R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, "Two-view contactlessfingerprint acquisition systems: a case study for clay artworks", Proc. of the2012 IEEE Workshop on Biometric Measurements and Systems for Security andMedical Applications (BioMS 2012), Salerno, Italy, September 14, 2012, pp. 1-8.
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition
Conclusions
• Touchless fingerprint recognition:
o Systems based on two-dimensional samples can be used in low-cost applications, but the samples present distortions
o Systems based on three-dimensional samples can obtain comparable accuracy with respect to traditional systems
o Touchless systems are characterized by higher usability, user acceptance, security.
o Touchless systems are partially compatible with AFIS
45
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition 46
Publications (1/2)
• Research booko A. Genovese, V. Piuri, and F. Scotti, Touchless Palmprint Recognition Systems", ser. Advances in
Information Security, vol. 60, S. Jajodia (ed.), Springer International Publishing, September 2014.
• Refereed Papers in International Journalso R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Toward unconstrained fingerprint recognition: a
fully-touchless 3-D system based on two views on the move", in IEEE Transactions on Systems, Manand Cybernetics: Systems, 2015.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Touchless fingerprint biometrics: a survey on2D and 3D technologies", in Journal of Internet Technology, vol. 15, no. 3, May 2014, pp. 325-332.
• Refereed Papers in Proceedings of InternationalConferences and Workshops
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Accurate 3D fingerprint virtual environment forbiometric technology evaluations and experiment design", in Proc. of the 2013 IEEE InternationalConference on Computational Intelligence and Virtual Environments for Measurement Systems andApplications (CIVEMSA 2013), Milan, Italy, July 15-17, 2013, pp. 43-48.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Contactless fingerprint recognition: a neuralapproach for perspective and rotation effects reduction", in Proc. of the 2013 IEEE Symposium onComputational Intelligence in Biometrics and Identity Management (CIBIM 2013), Singapore, April16-19, 2013, pp. 22-30.
© 2015 Angelo Genovese – Touchless and less-constrained 3D fingerprint recognition 47
Publications (2/2)
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Two-view contactless fingerprint acquisitionsystems: a case study for clay artworks", in Proc. of the 2012 IEEE Workshop on BiometricMeasurements and Systems for Security and Medical Applications (BioMS 2012), Salerno, Italy,September 14, 2012, pp. 1-8.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Virtual environment for 3-D syntheticfingerprints", in Proc. of the IEEE International Conference on Virtual Environments, Human-Computer Interfaces and Measurement Systems (VECIMS 2012), Tianjin, China, July 2-4, 2012, pp.48-53.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Quality measurement of unwrapped three-dimensional fingerprints: a neural networks approach", in Proc. of the 2012 International JointConference on Neural Networks (IJCNN 2012), Brisbane, Australia, June 10-15, 2012, pp. 1-8.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Fast 3-D fingertip reconstruction using a singletwo-view structured light acquisition", in Proc. of the 2011 IEEE Workshop on BiometricMeasurements and Systems for Security and Medical Applications (BioMS 2011), Milan, Italy,September 28, 2011, pp. 1-8.
o R. Donida Labati, A. Genovese, V. Piuri, and F. Scotti, Measurement of the principal singular point incontact and contactless fingerprint images by using computational intelligence techniques", in Proc.of the IEEE International Conference on Computational Intelligence for Measurement Systems andApplications (CIMSA 2010), Taranto, Italy, September 6-8, 2010, pp. 18-23.