This research website is aimed at providing a benchmark for the scientific community working on automatic detection of repeatable and distinctive keypoints in 3D data. 3D keypoint detection is currently a key stage for several 3D computer vision applications such as surface registration/alignment, SLAM and 3D camera pose estimation, 3D object recognition, surface matching.

This website includes a public dataset repository for testing 3D keypoint detectors, as well as experimental results concerning the evaluation proposed in [1,2]. Moreover, it provides a tool to compare newly developed algorithms with the state of the art.

References

[1] S. Salti, F. Tombari, L. Di Stefano, “A Performance Evaluation of 3D Keypoint  Detectors”, 3D Imaging, Modeling, Processing, Visualization and Transmission Conference (3DIMPVT), 2011

[2] F. Tombari, S. Salti, L. Di Stefano, “Performance Evaluation of 3D Keypoint Detectors”, International Journal of Computer Vision (IJCV), DOI: 10.1007/s11263-012-0545-4, to appear