Information Technology Reference
In-Depth Information
Table 5.
Comparison of four distribution graphes
palmprint
QEPD
finger
QEPD
palmprint
CSID
finger
CSID
Peak of inner-
class
0.10
0.12
0.09
0.14
Peak of outer-
class
0.24
0.34
0.21
0.32
Difference
0.14
0.22
0.12
0.18
ROC curve for 4 distances
100
95
CSID finger
CSID palmprint
QEPD finger
QEPD palmprint
90
85
80
75
70
65
60
55
0
10
20
30
40
50
60
70
80
90
FAR(%)
Fig. 4. ROC curve for 4 distances
The performance of a verification approach is measured by False Accept Rate
(FAR) and Correct Accept Rate (CAR). The ROC curve for all four distances
is shown in Figure 4. From the ROC curves for 4 distances, the performance
of finger QEPD, finger CSID, palmprint QEPD, and palmprint CISD size down
in turn. Therefore, we can safely conclude that QEPD outperforms CSID and
finger texture is a better discriminative biometrics than traditional biometric
palmprint for QEPD or CSID. This ROC curve has proven the inference above.
 
Search WWH ::




Custom Search