36
Feature-Preserving Reconstruction of Singular Surfaces with Tamal Dey, Qichao Que, Issam Safa, Lei Wang, Yusu Wang Computer science and Engineering The Ohio State University Xiaoyin Ge

Feature-Preserving Reconstruction of Singular Surfaces

  • Upload
    louise

  • View
    43

  • Download
    0

Embed Size (px)

DESCRIPTION

with Tamal Dey , Qichao Que , Issam Safa , Lei Wang, Yusu Wang Computer science and Engineering The Ohio State University. Feature-Preserving Reconstruction of Singular Surfaces. Xiaoyin Ge. Problem statement. Surface reconstruction of singular surface. input. output. - PowerPoint PPT Presentation

Citation preview

Page 1: Feature-Preserving Reconstruction of Singular Surfaces

Feature-Preserving Reconstruction of Singular Surfaces

with Tamal Dey, Qichao Que, Issam Safa, Lei Wang, Yusu WangComputer science and Engineering

The Ohio State University

Xiaoyin Ge

Page 2: Feature-Preserving Reconstruction of Singular Surfaces

Problem statement

Surface reconstruction of singular surface

input output

Page 3: Feature-Preserving Reconstruction of Singular Surfaces

Problem statement

Singular surface

A collection of smooth surface patches with boundaries.

glueintersectboundary

Page 4: Feature-Preserving Reconstruction of Singular Surfaces

Motivation and Previous work

2D manifold reconstruction

[AB99] Surface reconstruction by Voronoi filtering. AMENTA N., BERN M.

[ACDL02] A simple algorithm for homeomorphic surface reconstruction. AMENTA N., et. al.

[BC02] Smooth surface reconstruction via natural neighbor interpolation of distance functions. BOISSONNAT et. Al

[ABCO01] Point set surfaces. ALEXA et. al. …

Page 5: Feature-Preserving Reconstruction of Singular Surfaces

Motivation and Previous work

Feature aware method

[LCOL07] Data dependent MLS for faithful surface approximation. LIPMAN , et. al.

[ÖGG09] Feature preserving point set surfaces based on non-linear kernel regression, ÖZTIRELI, et.al

[CG06] Delaunay triangulation based surface reconstruction, CAZALS, et.al [FCOS05] Robust moving least-squares fitting with sharp features,

FLEISHMAN, et.al …

Page 6: Feature-Preserving Reconstruction of Singular Surfaces

Motivation

Need a simple yet effective reconstruction algorithm for all three singular surfaces.

Page 7: Feature-Preserving Reconstruction of Singular Surfaces

Our method: outline

Identify feature points

Reconstruct feature curves

Reconstruct singular surface

Page 8: Feature-Preserving Reconstruction of Singular Surfaces

Our method: outline

Identify feature points

Reconstruct feature curves

Reconstruct singular surface

Page 9: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

Gaussian-weighted graph Laplacian ( [BN02], Belkin-Niyogi, 2002)

Page 10: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

Gaussian-weighted graph Laplacian ([BQWZ12])

Gaussian kernel

Position difference

Page 11: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

Gaussian-weighted graph Laplacian, scaling ([BQWZ12])

boundarylow high

Page 12: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

surf B

surf A

intersectionlow high

Gaussian-weighted graph Laplacian, scaling ([BQWZ12])

Page 13: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

surf

A

surf B

glue (sharp feature)low high

Gaussian-weighted graph Laplacian, scaling ([BQWZ12])

Page 14: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

surf

A

surf B

Gaussian-weighted graph Laplacian (scaling, [BQWZ12])

boundary

surf B

surf A

intersection sharp feature

Page 15: Feature-Preserving Reconstruction of Singular Surfaces

Gaussian-weighted graph Laplacian

(I) Identify feature point

highlow

Page 16: Feature-Preserving Reconstruction of Singular Surfaces

(I) Identify feature point

Gaussian-weighted graph Laplacian Advantage:

Simple Unified approach Robust to noise

Page 17: Feature-Preserving Reconstruction of Singular Surfaces

Our method: outline

Identify feature points

Reconstruct feature curves

Reconstruct singular surface

Page 18: Feature-Preserving Reconstruction of Singular Surfaces

(II) reconstruct feature curve

Graph method proposed by [GSBW11]

[ Data skeletonization via reeb graphs, Ge, et.al , 2011]

Page 19: Feature-Preserving Reconstruction of Singular Surfaces

(II) reconstruct feature curve

Reeb graph ( from Rips-complex [DW11] )

Rips complexReeb graph(abstract)

Reeb graph(augmented)

Page 20: Feature-Preserving Reconstruction of Singular Surfaces

(II) reconstruct feature curve

Reeb graph a noisy graph

feature points Reeb graph

Page 21: Feature-Preserving Reconstruction of Singular Surfaces

(II) reconstruct feature curve

Graph simplification(denoise)

a zigzag graph

Page 22: Feature-Preserving Reconstruction of Singular Surfaces

Graph smoothening [KWT88] Use snake to smooth out the graph

(II) reconstruct feature curve

graph energy

graph Laplacian

Page 23: Feature-Preserving Reconstruction of Singular Surfaces

Graph smoothening Use snake to smoothen graph

graph Laplacian

(II) reconstruct feature curve

graph energy

align along feature

min( )

smoothen graph

Page 24: Feature-Preserving Reconstruction of Singular Surfaces

(II) reconstruct feature curve

Graph smoothening Use snake to smooth out the graph

Page 25: Feature-Preserving Reconstruction of Singular Surfaces

Our method: outline

Identify feature points

Reconstruct feature curves

Reconstruct singular surface

Page 26: Feature-Preserving Reconstruction of Singular Surfaces

(III) Reconstruct singular surface

Reconstruction [CDR07][CDL07]

[CDL07] A Practical Delaunay Meshing Algorithm for a Large Class of Domains, Cheng, et.al

[CDR07] Delaunay Refinement for Piecewise Smooth Complexes, Cheng-Dey-Ramos, 2007

Page 27: Feature-Preserving Reconstruction of Singular Surfaces

(III) Reconstruct singular surface

Weighted cocone

cocone weighted Delaunay[ACDL00] A simple algorithm for homeomorphic surface reconstruction, Amenta,-Choi-Dey -Leekha

Page 28: Feature-Preserving Reconstruction of Singular Surfaces

(III) Reconstruct singular surface

Weighted cocone

un-weighted point

weighted point

Page 29: Feature-Preserving Reconstruction of Singular Surfaces

(III) Reconstruct singular surface

Reconstruction Voronoi cell size weight∝ Give higher weight to points on the feature curve

Page 30: Feature-Preserving Reconstruction of Singular Surfaces

Experiment results

ab

cd

a. Octaflower 107K

b. Fandisk 114K

c. SphCube 65K

d. Beetle 63K

Page 31: Feature-Preserving Reconstruction of Singular Surfaces

Experiment results

Robust to noise

input with 1% noise result

Page 32: Feature-Preserving Reconstruction of Singular Surfaces

Experiment results

Perform much better than un-weighted cocone

Cocone Our method

Page 33: Feature-Preserving Reconstruction of Singular Surfaces

Conclusion and Future work

Conclusion Unified and simple method to handle all three

types of singular surfaces Robust to noise

Future work More robust system for real data Concave corner

Page 34: Feature-Preserving Reconstruction of Singular Surfaces

Acknowledgement

We thank all people who have helped us to demonstrate this method !

Most of the models used in this paper are courtesy of

AIM@SHAPE Shape Repository. The authors acknowledge the support of NSF under grants CCF-1048983, CCF-1116258 and CCF-0915996.

Page 35: Feature-Preserving Reconstruction of Singular Surfaces

Thank you!

Page 36: Feature-Preserving Reconstruction of Singular Surfaces

Conclusion and Future work

Real scanned data