![](bg6.png)
Table 1: Average computation time and percentage for
each step of ORB and DARP methods.
ORB DARP
ms % ms %
Keypoint detection 16.11 80.89 2.63 9.40
Normal estimation – – 14.99 53.56
Patch rectification – – 8.40 30.01
Orientation estimation 0.14 0.71 0.20 0.72
Patch description 3.67 18.40 1.77 6.31
Total 19.92 100.00 27.99 100.00
4 CONCLUSIONS
The DARP method has been introduced, which
exploits depth information to improve keypoint
matching. This is done by rectifying the patches
using the 3D information in order to remove
perspective distortions. The depth information is
also used to obtain a scale invariant representation of
the patches. It was shown that DARP can be used
together with existing keypoint matching methods in
order to help them to handle situations such as
oblique poses with respect to the viewing direction.
It supports both planar and non-planar objects and is
able to run in real-time.
As future work, tests with other keypoint
detectors and patch descriptors will be done.
Optimizations on normal estimation and patch
rectification are also planned, since they showed to
be the most time demanding steps of the technique.
REFERENCES
Berkmann, J., Caelli, T., 1994. Computation of surface
geometry and segmentation using covariance
techniques. In IEEE Transactions on Pattern Analysis
and Machine Intelligence, volume 16, issue 11, pages
1114–1116.
Del Bimbo, A., Franco, F., Pernici, F., 2010. Local
homography estimation using keypoint descriptors. In
WIAMIS’10, 11th International Workshop on Image
Analysis for Multimedia Interactive Services, 4 pages.
Eyjolfsdottir, E., Turk., M., 2011. Multisensory embedded
pose estimation. In WACV’11, IEEE Workshop on
Applications of Computer Vision, pages 23–30.
Hinterstoisser, S., Benhimane, S., Navab, N., Fua, P.,
Lepetit, V., 2008. Online learning of patch perspective
rectification for efficient object detection. In
CVPR’08, 21th IEEE Conference on Computer Vision
and Pattern Recognition, 8 pages.
Hinterstoisser, S., Kutter, O., Navab, N., Fua, P., Lepetit,
V., 2009. Real-time learning of accurate patch
rectification. In CVPR’09, 22th IEEE Conference on
Computer Vision and Pattern Recognition, pages
2945–2952.
Koser, K., Koch, R., 2007. Perspectively invariant normal
features. In ICCV’07, 11th IEEE International
Conference on Computer Vision, 8 pages.
Kurz, D., Benhimane, S., 2011. Gravity-aware handheld
augmented reality. In ISMAR’11, 10th IEEE
International Symposium on Mixed and Augmented
Reality, pages 111–120.
Lai, K., Bo, L., Ren, X., Fox, D., 2011. A large-scale
hierarchical multi-view RGB-D object dataset. In
ICRA’11, IEEE International Conference on Robotics
and Automation, pages 1817–1824.
Marcon, M., Frigerio, E., Sarti, A., Tubaro, S., 2012. 3D
wide baseline correspondences using depth-maps. In
Signal Processing: Image Communication, volume 27,
issue 8, pages 849–855.
Mikolajczyk, K., Tuytelaars, T., Schmid, C., Zisserman,
A., Matas, J., Schaffalitzky, F., Kadir, T., Van Gool,
L., 2005. A comparison of affine region detectors. In
International Journal of Computer Vision, volume 5,
issue 1–2, pages 43–72.
Morel, J., Yu, G., 2009. ASIFT: A new framework for
fully affine invariant image comparison. In SIAM
Journal on Imaging Sciences, volume 2, issue 2, pages
438–469.
Moreno-Noguer, F., Lepetit, V., Fua, P., 2007. Accurate
non-iterative O(n) solution to the PnP problem. In
ICCV’07, 11th IEEE International Conference on
Computer Vision, 8 pages.
Pagani, A., Stricker, D., 2009. Learning local patch
orientation with a cascade of sparse regressors. In
BMVC’09, 20th British Machine Vision Conference,
pages 86.1–86.11.
Rosten, E., Drummond, T., 2006. Machine learning for
high-speed corner detection. In ECCV’06, 9th
European Conference on Computer Vision, pages
430–443.
Rublee, E., Rabaud, V., Konolige, K., Bradski, G., 2011.
ORB: an efficient alternative to SIFT or SURF. In
ICCV’11, 15th IEEE International Conference on
Computer Vision, pages 2564–2571.
Wu, C., Clipp, B., Li, X., Frahm, J.-M., Pollefeys, M.,
2008. 3D model matching with viewpoint invariant
patches (VIPs). In CVPR’08, IEEE Conference on
Computer Vision and Pattern Recognition, 8 pages.
Yang, M., Cao, Y., Förstner, W., McDonald, J., 2010.
Robust wide baseline scene alignment based on 3d
viewpoint normalization. In ISVC’10, 6th
International Symposium on Visual Computing,
Lecture Notes in Computer Science, volume 6453,
pages 654–665.
VISAPP2013-InternationalConferenceonComputerVisionTheoryandApplications
656