Research Article
Efficient Iris Recognition Based on Optimal Subfeature Selection and Weighted Subregion Fusion
Table 4
Comparison of the computation complexity.
| Methods | Feature extraction (ms) | Matching (ms) | Feature extraction + Matching (ms) | Others |
| Daugman [3] | 682.5* | 4.3* | 686.8* | — | Wildes et al. [4] | 210.0* | 401.0* | 611.0* | Registration |
Boles and Boashash [5] | 170.3* | 11.0* | 181.3* | — | Ma et al. [6] | 260.2* | 8.7* | 268.9* | Feature reduction | Ma et al. [7] | 244.2* | 6.5* | 250.7* | — |
Roy and Bhattacharya [46] | 80.3∧ | 167.2∧ | 247.5∧ | Feature reduction |
Roy and Bhattacharya [47] | 20.3∧ | 130.4∧ | 150.7∧ | Feature reduction | Proposed (CASIA-V3 Interval) | 89.6 | 48.7 | 138.3 | Feature reduction & sub-region fusion | Proposed (CASIA-V3 Lamp) | 96.7 | 53.0 | 149.7 | Feature reduction & sub-region fusion |
|
|
Note. Data labeled by *come from Ma et al. [7] and ∧come from Roy and Bhattacharya [47].
|