© Jain, 2004
Fingerprint EnhancementFingerprint Enhancement
• To address the problem of poor quality fingerprints
Noisy image Enhanced image
© Jain, 2004
Motivation for Fingerprint
Enhancement
Minutia extracted from a good quality input image
Minutia extracted from a poor quality input image
© Jain, 2004
False MinutiaeFalse Minutiae
Quality Index = 0.04
False Minutiae=27
Quality Index = 0.53
False Minutiae=7
Quality Index = 0.96
False Minutiae=0
© Jain, 2004
Difficulty in Orientation Field
Estimation
© Jain, 2004
Fingerprint Regions
Recoverable
corrupted region
Unrecoverable
corrupted region
Well defined region
© Jain, 2004
Fingerprint Enhancement
Algorithm
• A bank of even-symmetric Gabor filters is applied
to the fingerprint image
• A ridge extraction algorithm is applied to each of
the filtered images and the corresponding ridge
map is obtained
• A voting algorithm is used to generate a coarselevel
ridge map and the unrecoverable region
masks
• Orientation field estimation is applied to the
coarse-level ridge map and from the computed
orientation field enhanced images are obtained
© Jain, 2004
Fingerprint Enhancement
Algorithm
• Gabor filters are both frequency and orientation
selective
• A Gabor filter with the correct local orientation
and local frequency can remove noise while
preserving the true ridges
• The algorithm is computationally intensive and
therefore applied only to the template image
• A fast enhancement algorithm is applied to the
input images during live verification
© Jain, 2004
Enhancement Algorithm
© Jain, 2004
Fingerprint EnhancementFingerprint Enhancement
A set of 8 Gabor filters
with different
orientations is used
Ridge
Voting
© Jain, 2004
Fingerprint as Oriented TextureFingerprint as Oriented Texture
(c) Fourier spectrum of (a)(a) Ridges in local region (b) Ridge directions in (a)
The ridge pattern in a fingerprint may be viewed as an oriented
texture pattern having a fixed dominant spatial frequency and
orientation in a local neighborhood. The frequency is due to the
inter-ridge spacing present in the fingerprint and the
orientation is due to the ridge flow pattern
© Jain, 2004
Even-symmetric Gabor FiltersEven-symmetric Gabor Filters
0o orientation filter 45o orientation filter
© Jain, 2004
Results of Enhancement
Input image Coarse ridge map Unrecoverable region mask
Estimated orientation field Enhanced image Extracted minutiae
© Jain, 2004
Fingerprint EnhancementFingerprint Enhancement
Minutiae extraction before enhancement
Minutiae extraction after enhancement
© Jain, 2004
Fast Enhancement Algorithm
© Jain, 2004
Results of Fast Enhancement
Algorithm
© Jain, 2004
Performance with EnhancementPerformance with Enhancement
Quantitative evaluation of enhancement module