Fingerprint Image Enhancement

1
Fingerprint Image Enhancement

• Joshua Xavier Munoz-Ramos

2
Motivation

• Method for fingerprint image enhancement
• Ridge structure in fingerprint images are not
  always well defined therefore, enhancement
  algorithm, is necessary
• A critical step in automatic fingerprint matching
  is extracting minutiae from the input fingerprint
  images. However, the performance of a minutiae
  extraction relies on the quality of the images.

3
Background

• Two Important ridge characteristics
• Ridge ending
• Ridge bifurcation

4
Approach

• Original Grayscale fingerprint image
• Local histogram equalization
• Wiener Filtering
• Binarization and thinning
• Morphological and further filtering (Anisotropic
  Filter)
• Enhanced binary image
• http//fvs.sourceforge.net/C27_icpr2000.pdf
• Fingerprint Image Enhancement
• Algorithm and Performance Evaluation
• Lin Hong, Student Member, IEEE, Yifei Wan, and
  Anil Jain, Fellow, IEEE

5
(No Transcript)
6
Weiner filter

• W(n1,n2)u (v2-n2)/(v2) I(n1,n2)-u
• 3x3 matrix
• Binary Thresholding (if I(n1,n2) gt local mean set
  to 1 other wise set to 0)

7
Weiner filter / binary
8
Morphing/ anisotropic filter

• Connecting the ridges through orientation fields

9
Anisotropic filter
Instead of using local gradients as a means of
controlling the anisotropism of filters, it uses
both a local intensity orientation and an
anisotropic measure to control the shape of the
filter. K(x0,x) exp-((x-x0) n ) 2/sig(x0)2
((x-x0) n(ortho))2/sig(x0)2 h(x0,x)
-2 10k(x0,x)
10
Results

• 71.1 percent FAR (using verification system )
  but only tested two fingerprints with 10
  different pics 7/10 were identified
• False ridges endings and bifurcations
• Need to test more fingerprints
• A good image has around 40 to 50 correct ridge
  endings and bifurcations
• (different method is to apply a garbor filter )
• Fingerprint acceptance rate
• (enhancement did not work as well as expected)
• Picture was clearer to see after enhancement, and
  the filters did smooth out noise
• However many false ridges and bifurcations
• Many parts where the picture was not clear my
  enhancement did not work.
• Future work. Fix the orientation field and the
  anisotropic filter.. Many details were lost.
• (citation) http//fvs.sourceforge.net/C27_icpr2000
  .pdf