Upright orientation of 3D shapes via tensor rank minimization

다국어 초록 (Multilingual Abstract)

In general, the upright orientation of a model is beneficial for human to recognize this model and is widely used in geometry processingand computer graphics. However, the orientation of the model obtained by existing technologies, such as 3D scanning systems ormodeling, may be far away from the right orientation. In order to solve this problem, a robust and efficient upright method is needed. Weobserve that when the model is aligned with the three axes, the rank of the three-order tensor constructed by this model is the lowest usually.

Inspired by this observation, we formulate the alignment of the 3D model with axes as a low-rank tensor optimization problemwhich is a global and unsupervised method and the genetic algorithm (GA) is used to solve this optimization problem. After the 3Dmodel has been aligned with the three axes, some geometric properties can be used to pick out the best upright orientation from the sixcandidate supporting bases easily. The three-order tensor is constructed by voxelizing the bounding box of the 3D model, and then fillingthe voxel element with zero or one based on whether it contains the points of the model or not. The experimental results demonstrate thatour method is robust, efficient and effective for all kinds of the models (manifold or non-manifold, man-made or non-artificial, or pointcloud).

참고문헌 (Reference)

1 H. B. Fu, "Upright orientation of man-made objects" 27 (27): 42-, 2008

2 Y. Jin, "Unsupervised upright orientation of man-made models" 74 (74): 99-108, 2012

3 H. Yamauchi, "Towards stable and salient multiview representation of 3D shapes" 40-45, 2006

4 P. Shilane, "The princeton shape benchmark" 167-178, 2004

5 T. G. Kolda, "Tensor decompositions and applications" 51 (51): 455-500, 2009

6 J. Liu, "Tensor completion for estimating missing values in visual data" 94 (94): 208-220, 2009

7 Z. D. Zhang, "TILT: transform invariant low-rank textures" 99 (99): 1-24, 2012

8 S. Tong, "Support vector machine active learning for image retrieval" 107-118, 2001

9 Dong Woo Lee, "Shape design of a tire contour based on approximation model" 대한기계학회 25 (25): 149-155, 2011

10 K. Xu, "Photo-inspired model-driven 3D object modeling" 30 (30): 80-, 2011

11 K. Xu, "Partial intrinsic reflection symmetry of 3D shapes" 28 (28): 138-, 2009

12 A. D. Belegundu, "Optimization Concepts and Applications in Engineering" Prentice Hall 1999

13 J. L. Wu, "Mesh saliency with global rarity" 2013

14 L. Guo, "Hole-filling by rank sparsity tensor decomposition for medical imaging" 94 (94): 396-399, 2011

15 P. Shilane, "Distinctive regions of 3D surfaces" 26 (26): 2007

16 E. Paquet, "Description of shape information for 2D and 3D objects" 16 (16): 103-122, 2000

17 F. R. Bach, "Consistency of trace norm minimization" (9) : 1019-1048, 2008

18 P. P. Vázquez, "Automatic view selection using viewpoint entropy and its application to image-based modelling" 22 (22): 689-700, 2003

19 J. B. Luo, "Automatic image orientation detection via confidence-based integration of low-level and semantic cues" 27 (27): 715-726, 2005

20 K. Hu, "Automatic generation of canonical views for CAD models" 17-24, 2011

21 J. W. H. Tangelder, "A survey of content based 3D shape retrieval methods" 39 (39): 441-471, 2008

22 J. Podolak, "A planar-reflective symmetry transform for 3D shapes" 25 (25): 549-559, 2006

23 P. J. Besl, "A method for registration of 3D shapes" 14 (14): 239-256, 1992

24 S. Biasotti, "3D shape matching through topological structures" 194-203, 2003

25 M. Ankerst, "3D shape histograms for similarity search and classification in spatial database" 207-226, 1999

26 "3D Scanning System"