Table 5 Registration results for scenario 3
Scenario 3 | Recall\(_C\) [%] | \(S_{Vg}\) | \(S_{Vs}\) | RMSD | |
|---|---|---|---|---|---|
Ground truth | 100 | 0.508 | 0.457 | 0.141 | |
Feature-based | SIFT + FPFH | 10.8 | 0.972 | 0.965 | 4.713 |
SIFT + PFHRGB | 10.5 | 0.974 | 0.966 | 4.746 | |
SIFT + RIFT | 9.7 | 0.975 | 0.969 | 4.751 | |
Deep learning | DCP | 6.4 | 0.985 | 0.979 | 4.842 |
PointNetLK | 8.8 | 0.977 | 0.972 | 4.771 | |
DeepGMR | 37.6 | 0.897 | 0.870 | 3.893 | |
GeoTransformer | 75.6 | 0.792 | 0.735 | 2.223 | |
Predator | 84.6 | 0.741 | 0.683 | 1.844 | |
Proposed solution | 100 | 0.508 | 0.591 | 0.152 | |