Point Cloud Upsampling and Normal Estimation using Deep Learning for Robust Surface Reconstruction

Rajat Sharma, Tobias Schwandt, Christian Kunert, Steffen Urban, Wolfgang Broll

Abstract

The reconstruction of real-world surfaces is on high demand in various applications. Most existing reconstruction approaches apply 3D scanners for creating point clouds which are generally sparse and of low density. These points clouds will be triangulated and used for visualization in combination with surface normals estimated by geometrical approaches. However, the quality of the reconstruction depends on the density of the point cloud and the estimation of the surface normals. In this paper, we present a novel deep learning architecture for point cloud upsampling that enables subsequent stable and smooth surface reconstruction. A noisy point cloud of low density with corresponding point normals is used to estimate a point cloud with higher density and appendant point normals. To this end, we propose a compound loss function that encourages the network to estimate points that lie on a surface including normals accurately predicting the orientation of the surface. Our results show the benefit of estimating normals together with point positions. The resulting point cloud is smoother, more complete, and the final surface reconstruction is much closer to ground truth.

Download


Paper Citation


in Harvard Style

Sharma R., Schwandt T., Kunert C., Urban S. and Broll W. (2021). Point Cloud Upsampling and Normal Estimation using Deep Learning for Robust Surface Reconstruction.In Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, ISBN 978-989-758-488-6, pages 70-79. DOI: 10.5220/0010211600700079


in Bibtex Style

@conference{visapp21,
author={Rajat Sharma and Tobias Schwandt and Christian Kunert and Steffen Urban and Wolfgang Broll},
title={Point Cloud Upsampling and Normal Estimation using Deep Learning for Robust Surface Reconstruction},
booktitle={Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,},
year={2021},
pages={70-79},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010211600700079},
isbn={978-989-758-488-6},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,
TI - Point Cloud Upsampling and Normal Estimation using Deep Learning for Robust Surface Reconstruction
SN - 978-989-758-488-6
AU - Sharma R.
AU - Schwandt T.
AU - Kunert C.
AU - Urban S.
AU - Broll W.
PY - 2021
SP - 70
EP - 79
DO - 10.5220/0010211600700079