Fingerprint Features

1

• Fingerprint Features

2

• 1 ) Introduction
• 2 ) Physiology
• 3 ) Uniqueness of a fingerprint configuration
• 4 ) Feature Extraction
• 5 ) Performance

3
1 ) Introduction

4
1 ) Introduction

• Most of fingerprint identification systems (like
  AFIS)?
• rely on minutiae (Level 12) only. While this
  information
• is sufficient for matching fingerprints in small
  databases,
• it is not discriminatory enough to provide good
  results
• on large collections of fingerprint images.
• M. Ray, P. Meenen, R. Adhami - A Novel
  Approach to Fingerprint Pore Extraction, IEEE,
  Mar. 2005
• AFIS...Automatic Fingerprint Identification
  System

5
1 ) Introduction fragment of 2 different
Fingerprints

• both show a bifurcation at the same location
• Examination based on Level 12 features match
• In combination with Level 3 features
• (e.g. relative pore position) no match

6
2 ) Physiology Fingerprint formation

• Fingerprints begin forming on the fetus 13th week
  of devellopment
• Bumps or ridge units are fusing together as they
  grow forming ridges
• Each ridge unit contains a pore which originates
  from a sweat gland from the dermis
• Pores are only found on ridges not in valleys
• sweat gland...Schweissdrüse

7
2 ) Physiology Some facts

• typical fingerprint 150 ridges
• A ridge 5 mm long contains appr. 10 ridge units
• Ridge width 0.5 mm
• Average number of pores / cm ridge 9-18 pores
• Pores do not disappear, move or generate over
  time
• Ashbaugh, D., Quantitative-Qualitative
  Friction Ridge Analysis, 1999, CRC Press
• Locard, Les pores et l'identification
  des criminals, Biologica, vol.2, pp. 257-365,
  1912

8
3 ) Uniqueness of a fingerprint configuration

• Ashbaugh model (1982)
• Assumptions
• Ridge units occur regularly along a ridge
• Position of a pore on a ridge unit is a random
  variable
• Independence between ridge units

9
3 ) Uniqueness of a fingerprint configuration

• Ashbaugh model (1982)
• 5 general areas where a pore may
• occur on the ridge unit
• Under the assumption of independence
• of ridge units
• P(pore in A)P(pore in B)...P(pore in E)
  Pp 0.2
• P(a sequence of N intra-ridge pores)PpN
  0.2 N
• P(a sequence of 20 intra-ridge pores)
  1.05 x 10-14

10
3 ) Uniqueness of a fingerprint configuration

• Rody and Stosz (1999)
• Estimated uniqueness of a
• sequence of intra-ridge pores
• based on measurements of real
• fingerprints (3748 distance
• measures)?
• Most common distance
• 13 pixels (0.3 mm)?

11
3 ) Uniqueness of a fingerprint configuration

• Rody and Stosz (1999)
• Pmeasured(a sequence of 20
• intra-ridge pores) 0.20120
• 1.16 x 10-14
• Assuming typical pore diameter
• of 5 pixels (115.5µm) allowing a
• displacement of 3 pixels (69.3µm)?
• P(a sequence of 20 ridge
• independent pores)
• 5.186 x 10-8

12
4 ) Feature extraction Pore extraction

• A matter of resolution
• Same fingerprint at different image
  resolutions
• 380 ppi (Identix 200DFR)
  (b) 500 ppi (Cross Match ID500) (c) 1000
  ppi (Cross Match ID1000)?
• 250-300 ppi minimum resolution for level 1
  level 2 features
• 500 ppi FBI standard for AFIS
• 1000 ppi minimum for extracting level 3
  features

13
4 ) Feature extraction Pore extraction

• A matter of condition
• Open pores may erroneously
• be interpreted as ridge endings

14
4 ) Feature extraction Pore extraction

• A matter of condition
• Dry skin produces distortions in
• the image that may be interpreted as
• pores

15
4 ) Feature extraction Pore extraction

• Anil K. Jain, Yi Chen, Meltem Demirkus Pores
  and Ridges High Resolution Fingerprint matching
  using level 3 features, IEEE Transactions on
  pattern analysis and machine intelligence,
  Vol.29, No.1, Jan. 2007

16
4 ) Feature extraction Pore extraction

• Presence of pores is not guaranteed
• 2 images of the same finger for different skin
  conditions

17
4 ) Feature extraction Contour Extraction

• Wavelet Transform Gabor
  enhanced image Ridge Contours
• - Wavelet response

18
5 ) Performance

• Hierarchical matching
• Level 1 orientation field Level
  2 feature location Level 3 pores ridge
  contour

19
5 ) Performance

• Test database
• 1.640 fingerprint images
• (Crossmatch 1000ID Sensor)

20
Referenzen

• M. Ray, P. Meenen, R. Adhami - A Novel
  Approach to Fingerprint Pore Extraction, IEEE,
  Mar.
• 2005
• Ashbaugh, D., Quantitative-Qualitative Friction
  Ridge Analysis, 1999, CRC Press
• Locard, Les pores et l'identification des
  criminals, Biologica, vol.2, pp. 257-365, 1912
• Anil K. Jain, Pores and Ridges High
  Resolution Fingerprint matching using level 3
  features,
• IEEE ransactions on pattern analysis and
  machine intelligence, Vol.29, No.1, Jan. 2007

21

• Thanks for listening!