Research Article
An ROI Extraction Method of Finger Vein Images Based on Large Receptive Field Gradient Operator for Accurate Localization of Joint Cavity
Table 4
EERs calculated by different ROI extraction methods.
| Feature | ROI | MMCBNU_6000 (%) | UTFVP (%) | FV-USM (%) |
| Maximum curvature | Yao’s method | 6.33 | 1.39 | 8.85 | Single sliding window method | 6.76 | 4.24 | 5.89 | Fixed window | 4.97 | 3.29 | 2.47 | Dual-sliding windows | 5.77 | 4.51 | 6.38 | Single-line accumulation | 7.19 | 3.59 | 4.50 | Proposed method | 4.91 | 2.77 | 2.81 |
| Repeated line tracking | Yao’s method | 4.26 | 1.74 | 13.22 | Single sliding window method | 3.76 | 1.35 | 3.59 | Fixed window | 2.38 | 1.04 | 1.52 | Dual-sliding windows | 3.22 | 1.53 | 4.84 | Single-line accumulation | 3.95 | 1.12 | 3.34 | Proposed method | 3.73 | 0.97 | 1.69 |
|
|