The experiments were conducted on the high resolution partial
fingerprint images using the proposed pore extraction and pore matching
methods. The
implementation of the algorithm was done in MATLAB software. The Image
Processing Toolbox and the Computer Vision Toolbox were used.
DATABASE
For reliable extraction of pores a high resolution
fingerprint images of minimum 1000 ppi are required. Unfortunately, so far
there is no such high resolution fingerprint image database freely available in
the public domain except the database developed by National Institute of
Standards and Technology (NIST) called SD30 (Special Database) which is very
costly. Q. Zhao et al. [7] developed a high resolution partial fingerprint
image database [22] using a scanner. The image size is of 320 pixels in width
and 240 pixels in height. The image resolution was 1200 ppi.
Pore Detection Accuracy of DAPM
We first assess the pore detection accuracy of the proposed
two pore extraction model namely Dynamic Anisotropic Pore Model and Difference
of Gaussian model. For this purpose we choose a set of 10 fingerprint images
from the database. Fig.6.2.1 and Fig.6.2.2 shows the examples of pore
extraction result of the two pore extraction model. In addition to the visual
evaluation of the pore detection results, we calculated the pore detection
accuracy on the 10 fingerprint images by using the two metric: True Detection Rate (TDR) and False
Detection Rate (FDR).
A
good pore extraction should have a high TDR and a low FDR simultaneously. Table 6.2.1 and Table 6.2.2 list the
average detection accuracy of the two pore extraction considered as a true pore
whereas the one which lies outside the ridge i.e. on valley are considered as
false pores.
Table
4.2.1: Performance Measure of Dynamic Anisotropic Pore Model
|
Images
|
TDR
|
Percentage
|
FDR
|
Percentage
|
|
Image_1
|
54/473
|
11.41 %
|
335/473
|
67.67 %
|
|
Image_2
|
24/389
|
06.16 %
|
338/389
|
86.88 %
|
|
Image_3
|
69/448
|
15.40 %
|
318/448
|
70.98 %
|
|
Image_4
|
48/422
|
11.37 %
|
320/422
|
75.82 %
|
|
Image_5
|
68/456
|
14.91 %
|
364/456
|
79.82 %
|
|
Image_6
|
44/412
|
10.67 %
|
308/412
|
74.75 %
|
|
Image_7
|
48/491
|
09.77 %
|
353/491
|
71.89 %
|
|
Image_8
|
44/374
|
11.76 %
|
263/374
|
70.32 %
|
|
Image_9
|
32/363
|
08.81 %
|
272/363
|
74.93 %
|
|
Image_10
|
48/320
|
15.00 %
|
243/320
|
75.93 %
|
 |
| Fig1: Results of DAPM pore extraction; (a),(c),(e) and (g) are input images; (b), (d), (f) and (h) shows detected pores on input images respectively |
No comments:
Post a Comment