Finger Vein Recognition Based on (2D)² PCA and Metric Learning

Table 1

Methods for personal authentication using finger vein recognition.

References	Method	Database fingers × samples per each	Performance
[4]	Linetracking	339×2 images	EER: 0.145%
[5]	Mean curvature	125×9 images	EER: 0.25%
[13]	Wide line detector	10,140×5 images	EER: 0.87%
[14]	Statistical vein energy	100×10 images	CCR: 98.7%
[15]	Moment invariants	50×4 images	EER: 8.93%
[16]	Sliding window matching	76×6 images	EER: 0.54%
[10]	Manifold learning	164×70 images	EER: 0.8%
[7]	PCA + BP network	10×10 images	CCR: 99%
[11]	PCA + LDA + SVM	10×10 images	CCR: 98%