Professional Documents
Culture Documents
S. Tulyakov, F. Farooq and V. Govindaraju Center for Unified Biometrics and Sensors SUNY at Buffalo, New York, USA
It is impossible to learn the original password given stored hash value of it.
Hash space h
h(f1) h(f2)
Since match algorithm will work with the values of hash functions, similar fingerprints should have similar hash values rotation and translation of original image should not have big impact on hash values partial fingerprints should be matched
Existing Approaches
Davida, Frankel, Matt (1998) - use error correcting codes, features should be ordered Biometric encryption (Soutar et al., 1998) - use filters for Fourier transform of fingerprint image - translation is accounted for, but not rotation
Existing Approaches
Ratha, Connell, Bolle (2001) - polynomial transform; need alignment. Juels, Sudan (2002) - map points to the values of error correcting codes , introduce variation by adding some other points. Follow-ups: Clancy et al, 2003 Uludag, Jain (2004)
( x a ) ( y b )i x yi
( x yi ) (a bi ) ! ( x a) ( y b)i
z ! x yi
|z| y
N
x
r v z !| r | (cos N i sin N )v | z | (cos U i sin U ) !| r || z | (cos N cos U i sin N cos U i cos N sin U i 2 sin N sin U ) !| r || z | (cos N cos U sin N sin U i (sin N cos U cos N sin U )) !| r || z | (cos(U N ) i sin(U N ))
rz
| z |v| r |
rz t
r !| r | (cos N i sin N )
Multiplying by r means rotating by angle N and scaling by factor | r | .
z |z|
Transformation function
If (c1 , c2 , - , cn ) is a set of minutia points of first d 2 ,- d fingerprint and (c1 , cd , cn ) is a set of minutia points of second fingerprint (same finger), then we assume that there ! is a transformation f ( z ) ! rz t such that cid f (ci ) ! rci t for any i ! 1, - n .
f ( z ) ! rz t
The values of these symmetric functions do not depend on the order of minutia points.
d 2 ,- n ) d d h2 (c1 , cd , cd! c1 c2 - cd n ! (rc1 t ) 2 (rc2 t ) 2 - (rcn t ) 2 ! r 2 (c1 c2 - cn ) 2rt (c1 c2 - cn ) nt 2 ! r h2 (c1 , c2 ,- , cn ) 2rth1 (c1 , c2 ,- , cn ) nt
2 2 2 2 2
dd Thus hi (c1 , c2 , - , cdcan be expressed as a linear n) combinations of h j (c1 , c2 , - , cn ), j e i with coefficients depending on transformation parameters r and t.
Denote:
Thus
Verifying fingerprint match using hash functions When r and t are found we can use higher order hash functions to check if fingerprints match. For example, if extracted minutia set is identical to the stored in the database, then for the hash function of third order we should get: 3 2 2 3
d h3 ! r h3 3r th2 3rt h1 nt
The difference between two parts of above equation can serve as a confidence measure for matching two sets of minutia points.
g i (d1 , d 2 ,- , d n ) ! d d 2 - d n
i 1
The small changes in locations of minutia points result in big changes of symmetric functions of higher orders. Thus we limited ourselves to the symmetric functions of 1st and 2nd orders.
Goodness of Match
dist (hi , hid i , g id , t ) ! E1 | h1d rh1 nt | ,g ,r d E 2 | h2 r 2 h2 2rth1 nt 2 | d E 3 | g1 rg1 nt | d E 4 | g 2 r 2 g 2 2rtg1 nt 2 |
For all local subsets find how many subsets are matched and whether values of r and t are similar.
For each minutiae point, find the 3 nearest neighbors and form 3 triplets that always include the initial minutia.
Enrollment: 1. For each triplet generated let (c1,c2,c3) and (d1,d2,d3) be the locations and directions of the minutia 2. Compute hash functions h1 = (c1 + c2 + c3)/3
g1 = (d1 + d2 + d3)/3
h2 = (c12 + c22 + c32)/3
Experimental results
10 0
80
Co.3: 3 pts and 2 hash fns ERR = 3% Original no-hash matching ERR = 1.7% Co.2: 3 pts and 1 hash fn Co.1: 2 points and 1 hash fn tested on FVC2002 set, with 2800 genuine tests and 4950 impostor tests
0 2 4 6 8 10 12
60
40
20
Algorithm limitations
Different local minutia sets can have same hash value sets. Thus the expected performance of the algorithm is lower than the performance of the matching algorithm using all available fingerprint information. Usually there are less matching hash values than matching minutiae. This means bigger difficulty in producing good match score, and setting match thresholds. c2 d d c1 c2 c
1
h1 c3
h1d
d c3
Thank you !
References 1. Davide Maltoni, Dario Maio, Anil K. Jain and Salil Prabhakar, Handbook of Fingerprint Recognition, SpringerVerlag, New York, 2003 2. Colin Soutar, Danny Roberge, Alex Stoianov, Rene Gilroy and B.V.K. Vijaya Kumar, Biometric Encryption, in ICSA Guide to Cryptography, R.Nichols, ed. (McGraw-Hill, 1999) 3. G.I. Davida, Y. Frankel, and B.J. Matt. On enabling secure applications through offline biometric identification. In IEEE Symposium on Privacy and Security, 1998. 4. Tsai-Yang Jea, Viraj S. Chavan, Venu Govindaraju and John K. Schneider, Security and matching of partial fingerprint recognition systems, In SPIE Defense and Security Symposium, 2004.
Security
If the number of stored hash functions
j
hi is less than
the number of minutia points, it is not possible to find the positions of minutia points c from local hash values. Using system of hash equations is difficult, since it is not known which minutia correspond to particular hash value.
x x x x
o
x
(a)
x x
x
(b)
o
x
x x
o
x
(c)
o
x