An all-digital ring-wedge detector system is presented that
simulates the analog multielement array commonly used in coherent
optoelectronic processors. The system is applicable with either
hard-copy or digital imagery. Using neural-network software, we
demonstrate high accuracy for the recognition of fingerprints,
including both orientation and wide-scale size-independent sortings by
using ring-only and wedge-only input neurons, respectively. Also,
the system is applied on windowed subregions of fingerprint imagery,
providing a feature set that summarizes localized information about
spatial-frequency content and edge-angle correlations. Examples are
presented in which this localized spatial-frequency information is used
to produce local ridge-orientation maps and to detect regions of poor
print quality. In summary, both direct-image data and
spatial-transform data are found to be important.
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Fingerprint-Sorting Accuracy for Both Ring and Wedge Data
from Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints
(20 from Each Person) in the Testing Set with 2
Errorsa
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
19
1
F3
20
F4
20
F5
20
F6
19
1
F7
20
F8
20
For example, as shown 20 prints from
person F1 are classified correctly by neuron 1.
Table 2
Fingerprint-Sorting Accuracy for Only Ring Data from
Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 10 Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
18
1
1
F2
19
1
F3
20
F4
1
19
F5
20
F6
19
F7
4
16
F8
1
1
18
Table 3
Fingerprint-Sorting Accuracy for Only Wedge Data from
Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 7 Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
19
1
F2
19
1
F3
19
1
F4
20
F5
19
1
F6
19
1
F7
20
F8
2
18
Table 4
Fingerprint-Sorting Accuracy for Binarized Print Imagery
Data for a Data Set of 160 Separate Thumbprints in the Testing Set
with Zero Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
20
F3
20
F4
20
F5
20
F6
20
F7
20
F8
20
Table 5
Fingerprint-Sorting Accuracy for Only Ring Data from
Binarized Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 1 Error
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
20
F3
20
F4
20
F5
20
F6
20
F7
20
F8
1
19
Table 6
Fingerprint-Sorting Accuracy for Only Wedge Data from
Binarized Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 1 Error
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
19
1
F3
20
F4
20
F5
20
F6
20
F7
20
F8
20
Table 7
Fingerprint-Sorting Accuracy for Binarized Print Imagery
for a Data Set of 160 Separate Thumbprints in the Testing Set as Well
as 160 Print Images from Unknown Persons with 0 Errors, 0 False
Accepts, and 4 False Rejects
Testing Set
Maximum-Valued Output Neuron
Reject
1
2
3
4
5
6
7
8
F1
18
2
F2
20
F3
20
F4
20
F5
20
F6
20
F7
19
1
F8
19
1
Unknown
160
Table 8
Fingerprint-Verification Results for 8 Independent Trials,
Each Using Binarized Imagery from Live Scanned Fingerprints with an
Overall False-Accept Rate of 0/160 and an Overall False-Reject Rate
of 5/160
Person
Errors
False Reject
False Accept
F1
0 in 20
0 in 160
F2
0 in 20
0 in 160
F3
2 in 20
0 in 160
F4
0 in 20
0 in 160
F5
0 in 20
0 in 160
F6
1 in 20
0 in 160
F7
2 in 20
0 in 160
F8
0 in 20
0 in 160
Table 9
Summary of Results for Fingerprint Experiments for the
All-Digital Ring-Wedge Detector
Section in This Paper
Description of Experiment
Results of Experiment
Subsection 3.A
Gray-scale sorting
Good sorting results
2/160 Errors (ring and wedge data)
10/160 Errors (ring data only)
7/160 Errors (wedge data only)
Subsection 3.B
Binary sorting
Improved sorting results
0/160 Errors (ring and wedge data)
1/160 Errors (ring data only)
1/160 Errors (wedge data only)
Subsection 3.C
Sorting with reject category
Demonstrates excellent recognition
4/160 False rejects
0/160 False accepts
0/156 Sorting errors
Sections 4 and 5
Fingerprint verification
Demonstrates practical application
5/160 False rejects
0/160 False accepts
Section 6
Localized ring-wedge transform
Effectively combines both image-domain and spatial-transform-domain information
Useful for image quality assessment and characterizing local print characteristics
Tables (9)
Table 1
Fingerprint-Sorting Accuracy for Both Ring and Wedge Data
from Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints
(20 from Each Person) in the Testing Set with 2
Errorsa
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
19
1
F3
20
F4
20
F5
20
F6
19
1
F7
20
F8
20
For example, as shown 20 prints from
person F1 are classified correctly by neuron 1.
Table 2
Fingerprint-Sorting Accuracy for Only Ring Data from
Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 10 Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
18
1
1
F2
19
1
F3
20
F4
1
19
F5
20
F6
19
F7
4
16
F8
1
1
18
Table 3
Fingerprint-Sorting Accuracy for Only Wedge Data from
Gray-Scale Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 7 Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
19
1
F2
19
1
F3
19
1
F4
20
F5
19
1
F6
19
1
F7
20
F8
2
18
Table 4
Fingerprint-Sorting Accuracy for Binarized Print Imagery
Data for a Data Set of 160 Separate Thumbprints in the Testing Set
with Zero Errors
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
20
F3
20
F4
20
F5
20
F6
20
F7
20
F8
20
Table 5
Fingerprint-Sorting Accuracy for Only Ring Data from
Binarized Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 1 Error
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
20
F3
20
F4
20
F5
20
F6
20
F7
20
F8
1
19
Table 6
Fingerprint-Sorting Accuracy for Only Wedge Data from
Binarized Imagery for a Data Set of 160 Separate Thumbprints in the
Testing Set with 1 Error
Testing Set
Maximum-Valued Output Neuron
1
2
3
4
5
6
7
8
F1
20
F2
19
1
F3
20
F4
20
F5
20
F6
20
F7
20
F8
20
Table 7
Fingerprint-Sorting Accuracy for Binarized Print Imagery
for a Data Set of 160 Separate Thumbprints in the Testing Set as Well
as 160 Print Images from Unknown Persons with 0 Errors, 0 False
Accepts, and 4 False Rejects
Testing Set
Maximum-Valued Output Neuron
Reject
1
2
3
4
5
6
7
8
F1
18
2
F2
20
F3
20
F4
20
F5
20
F6
20
F7
19
1
F8
19
1
Unknown
160
Table 8
Fingerprint-Verification Results for 8 Independent Trials,
Each Using Binarized Imagery from Live Scanned Fingerprints with an
Overall False-Accept Rate of 0/160 and an Overall False-Reject Rate
of 5/160
Person
Errors
False Reject
False Accept
F1
0 in 20
0 in 160
F2
0 in 20
0 in 160
F3
2 in 20
0 in 160
F4
0 in 20
0 in 160
F5
0 in 20
0 in 160
F6
1 in 20
0 in 160
F7
2 in 20
0 in 160
F8
0 in 20
0 in 160
Table 9
Summary of Results for Fingerprint Experiments for the
All-Digital Ring-Wedge Detector
Section in This Paper
Description of Experiment
Results of Experiment
Subsection 3.A
Gray-scale sorting
Good sorting results
2/160 Errors (ring and wedge data)
10/160 Errors (ring data only)
7/160 Errors (wedge data only)
Subsection 3.B
Binary sorting
Improved sorting results
0/160 Errors (ring and wedge data)
1/160 Errors (ring data only)
1/160 Errors (wedge data only)
Subsection 3.C
Sorting with reject category
Demonstrates excellent recognition
4/160 False rejects
0/160 False accepts
0/156 Sorting errors
Sections 4 and 5
Fingerprint verification
Demonstrates practical application
5/160 False rejects
0/160 False accepts
Section 6
Localized ring-wedge transform
Effectively combines both image-domain and spatial-transform-domain information
Useful for image quality assessment and characterizing local print characteristics