Research Article
Encoding Coefficient Similarity-Based Multifeature Sparse Representation for Finger Vein Recognition
Table 1
Performance comparison between SFSR and SMSR methods.
| | | HKPU-FV | SDU-FV | Method | Feature | Recognition (%) | EER (%) | Recognition (%) | EER (%) |
| SFSR | Grayvalue | 91.67 | 2.24 | 72.38 | 7.72 | Gabor | 91.88 | 3.11 | 60.06 | 18.99 | LBP | 98.93 | 0.22 | 84.07 | 5.35 | HOG | 98.82 | 0.63 | 77.94 | 8.68 |
| Proposed SMSR | Grayvalue + Gabor | 98.08 | 0.73 | 78.67 | 9.95 | Grayvalue + HOG | 99.68 | 0.33 | 90.15 | 4.25 | Gabor + LBP | 99.68 | 0.32 | 91.72 | 3.44 | Gabor + HOG | 99.25 | 0.44 | 87.74 | 5.50 | LBP + HOG | 99.79 | 0.11 | 91.82 | 3.41 | Grayvalue + LBP | 99.89 | 0.12 | 93.87 | 2.61 |
|
|
Bold values represent the best results.
|