Department of Electrical Engineering (ISY)

Density Adaptive Point Set Registration

Felix Järemo Lawin, Martin Danelljan, Fahad Khan, Per-Erik Forssén, Michael Felsberg
CVPR18, Salt Lake City, Utah, USA
IEEE Conference on Computer Vision and Pattern Recognition
June 2018

Abstract

Probabilistic methods for point set registration have demonstrated competitive results in recent years. These techniques estimate a probability distribution model of the point clouds. While such a representation has shown promise, it is highly sensitive to variations in the density of 3D points. This fundamental problem is primarily caused by changes in the sensor location across point sets. We revisit the foundations of the probabilistic registration paradigm. Contrary to previous works, we model the underlying structure of the scene as a latent probability distribution, and thereby induce invariance to point set density changes. Both the probabilistic model of the scene and the registration parameters are inferred by minimizing the Kullback-Leibler divergence in an Expectation Maximization based framework. Our density-adaptive registration successfully handles severe density variations commonly encountered in terrestrial Lidar applications. We perform extensive experiments on several challenging real-world Lidar datasets. Our results clearly demonstrate that the proposed density-adaptive registration significantly outperforms state-of-the-art probabilistic methods.

Full Paper

Full paper is on ArXiv.
Open access paper on the CVF site [PDF] [suppl].
Source is available on GitHub.

---

Bibtex entry

@InProceedings{jaremo18a,
  author = 	 {Felix J\"aremo Lawin and Martin Danelljan and Fahad Khan and Per-Erik Forss\'en and Michael Felsberg},
  title = 	 {Density Adaptive Point Set Registration},
  booktitle = {{IEEE} Conference on Computer Vision and Pattern Recognition},
  year = 	 {2018},
  month = 	 {June},
  address = 	 {Salt Lake City, Utah, USA},
  publisher =    {Computer Vision Foundation},
  note = 	 {VR Project: Learnable Camera Motion Models, 2014-5928}
}