September 2019 Full - page 683

Table 9. The basic information of the patches in the aerial image data set. KeyPoint1-B and KeyPoint2-B are the numbers of
interest points extracted from an image pair using the BRISK algorithm, while KeyPoint1-S and KeyPoint2-S use SIFT algorithm.
Camera Resolution KeyPoints1-B KeyPoints2-B KeyPoints1-S KeyPoints2-S
Pair1 Leica RCD30 3000×2244 79 507 67 232 47 393 44 942
Pair2 Leica RCD30 3000×2244 67 232 56 685 44 942 39 684
Pair3 Leica RCD30 3000×2244 56 685 64 845 39 684 37 898
Table 10. The number of correspondences of the three image pairs using different combinations of detector and descriptors. PPD
is the proposed learned deep descriptor. Bold numbers indicate that the corresponding descriptors have the top performance.
BRISK BRISK+TFeat BRISK+PPD SIFT SIFT+TFeat SIFT+PPD
Pair1 163 263 462 157 188 342
Pair2 91 126 198 107 127 217
Pair3 18 34 74 19 22 59
Figure 17. Feature point matching results of image pair Pair1 after outlier removal using RANSAC through different
combinations of different detectors and descriptors.
Figure 18. Feature point matching results of image pair Pair2 removal using RANSAC through different combinations of
different detectors and descriptors.
Figure 19. Feature point matching results of image pair Pair3 after outlier removal using RANSAC through different
combinations of different detectors and descriptors.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING September 2019 683