Image Processing Reference
In-Depth Information
that of (20 × 60). However, in row direction experiments, in both cases ((10 × 20) and (20 × 60)
sample sizes), the accuracy rate goes upper than that of column direction method with less
time-consumption duration.
FIGURE 5
Accuracy rates obtained using K2DPCA in row direction on finger vein database:
(a) (10 × 20) sample size and (b) (20 × 60) sample size.
In this part, we give a summary of the whole experiments and their corresponding results
for the sake of a beter comparison. We have chosen the highest accuracies of each method in
all implementations and their corresponding dimension of feature vector. All the mentioned
information are indicated in
Table 1
in addition to the duration of time each algorithm con-
sumed to analyze the data. As
Table 1
shows, the maximum accuracy of the row direction ana-
lysis is higher than that of column direction in all the different experiments. Furthermore, it is
observed that not only it leads to higher accuracy, but also its dimension of feature vector is
much less than that of column method in all implementations implying that the row direction
method can be even faster than the column direction method in real-time system as it reaches
the higher accuracy using less feature vectors.
Table 1
Comparison of the Proposed Algorithm in Row and Column Direction
Images
to
Train
Max Ac-
curacy
(%)
Feature
Vector
No
Duration of
Experiments
(s)
Images
to
Train
Max Ac-
curacy
(%)
Feature
Vector
No
Method
(20 × 60)
Method
(10 × 20)
Column dir-
ection
analysis
2
91.63
60
1324.8372
Column dir-
ection
analysis
2
91.55
20
3
97
60
1676.553
3
95.14
20
4
97.83
15
1805.3797
4
97.83
20
Row direc-
tion ana-
lysis
2
95.75
7
223.73
Row direc-
tion ana-
lysis
2
95.75
10
3
97.71
5
311.2961
3
96.14
2
4
99.17
7
318.9148
4
98.17
3
7 Conclusion
In this paper, we have proposed a new method to enhance the performance of finger vein re-
cognition and analyzed two different aspects of applying it in order to determine the most ap-
propriate one. Our algorithm uses kernel mapping in two different directions to transfer the
input data to another space where applying 2DPCA merits the final output of the system. We
also used Euclidian distance as classifier in the last step of the algorithm. Extensive experi-
ments were conducted on our database using three different numbers of images for training.
Search WWH ::
Custom Search