This paper proposes a method for reconstructing accurate 3D surface points. To this end, robust and dense reconstruction with Shape-from-Silhouettes (SfS) and accurate multiview stereo are integrated. Unlike gradual shape shrinking and/or bruteforce large space search by existing space carving approaches, our method obtains 3D points by SfS and stereo independently, and then selects correct ones from them. The point selection is achieved in accordance with spatial consistency and smoothness of 3D point coordinates and normals. The globally optimized points are selected by graph-cuts. Experimental results with several subjects containing complex shapes demonstrate that our method outperforms existing approaches and our previous method.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Norimichi UKITA, Kazuki MATSUDA, "3D Reconstruction with Globally-Optimized Point Selection" in IEICE TRANSACTIONS on Information,
vol. E95-D, no. 12, pp. 3069-3077, December 2012, doi: 10.1587/transinf.E95.D.3069.
Abstract: This paper proposes a method for reconstructing accurate 3D surface points. To this end, robust and dense reconstruction with Shape-from-Silhouettes (SfS) and accurate multiview stereo are integrated. Unlike gradual shape shrinking and/or bruteforce large space search by existing space carving approaches, our method obtains 3D points by SfS and stereo independently, and then selects correct ones from them. The point selection is achieved in accordance with spatial consistency and smoothness of 3D point coordinates and normals. The globally optimized points are selected by graph-cuts. Experimental results with several subjects containing complex shapes demonstrate that our method outperforms existing approaches and our previous method.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E95.D.3069/_p
Copy
@ARTICLE{e95-d_12_3069,
author={Norimichi UKITA, Kazuki MATSUDA, },
journal={IEICE TRANSACTIONS on Information},
title={3D Reconstruction with Globally-Optimized Point Selection},
year={2012},
volume={E95-D},
number={12},
pages={3069-3077},
abstract={This paper proposes a method for reconstructing accurate 3D surface points. To this end, robust and dense reconstruction with Shape-from-Silhouettes (SfS) and accurate multiview stereo are integrated. Unlike gradual shape shrinking and/or bruteforce large space search by existing space carving approaches, our method obtains 3D points by SfS and stereo independently, and then selects correct ones from them. The point selection is achieved in accordance with spatial consistency and smoothness of 3D point coordinates and normals. The globally optimized points are selected by graph-cuts. Experimental results with several subjects containing complex shapes demonstrate that our method outperforms existing approaches and our previous method.},
keywords={},
doi={10.1587/transinf.E95.D.3069},
ISSN={1745-1361},
month={December},}
Copy
TY - JOUR
TI - 3D Reconstruction with Globally-Optimized Point Selection
T2 - IEICE TRANSACTIONS on Information
SP - 3069
EP - 3077
AU - Norimichi UKITA
AU - Kazuki MATSUDA
PY - 2012
DO - 10.1587/transinf.E95.D.3069
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E95-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 2012
AB - This paper proposes a method for reconstructing accurate 3D surface points. To this end, robust and dense reconstruction with Shape-from-Silhouettes (SfS) and accurate multiview stereo are integrated. Unlike gradual shape shrinking and/or bruteforce large space search by existing space carving approaches, our method obtains 3D points by SfS and stereo independently, and then selects correct ones from them. The point selection is achieved in accordance with spatial consistency and smoothness of 3D point coordinates and normals. The globally optimized points are selected by graph-cuts. Experimental results with several subjects containing complex shapes demonstrate that our method outperforms existing approaches and our previous method.
ER -