Research Article
Robust Detection Model of Vascular Landmarks for Retinal Image Registration: A Two-Stage Convolutional Neural Network
Table 4
Comparison of the detection performance on RetinaCheck database.
| Method | DRIVE | IOSTAR | Precision | Recall | F1 score | Precision | Recall | F1 score |
| Azzopardi and Petkov [47] | 0.400 | 0.740 | 0.520 | 0.630 | 0.330 | 0.430 | Abbasi-Sureshjani et al. [30] | 0.750 | 0.610 | 0.670 | 0.470 | 0.600 | 0.520 | Uslu and Bharath [48] | 0.650 | 0.690 | 0.670 | 0.520 | 0.670 | 0.590 | Zhao et al. [49] | 0.710 | 0.700 | 0.700 | 0.620 | 0.570 | 0.600 | Proposed method | 0.805 | 0.776 | 0.790 | 0.524 | 0.842 | 0.646 |
|
|