Table 1. The strut detection results of the presented method for each validation data set. For each data set, the number of frames containing struts (Frame No.), the numbers of struts in the ground truth (GT), the percentages of true positive (TP) and false positive (FP) are given.
| Strut status | Data set | Frame No. (with struts) | No. of GT | TP (%) | FP (%) |
|---|---|---|---|---|---|
| Baseline | 1 | 89 | 630 | 85.5 | 3.8 |
| 2 | 94 | 770 | 92.2 | 3.3 | |
| 3 |
93 |
783 |
92.5 |
2.0 |
|
| Follow-up | 4 | 97 | 868 | 93.7 | 0.0 |
| 5 | 91 | 819 | 98.0 | 0.6 | |
| 6 |
97 |
821 |
90.4 |
1.8 |
|
| Total | - | 561 | 4691 | 93.7 | 1.8 |