Curve fitting to point clouds - PowerPoint PPT Presentation

1 / 90
About This Presentation
Title:

Curve fitting to point clouds

Description:

Reconstructing B-spline curves from point clouds ... Shape Modeling and Applications, 2005 International Conference. CAGD&CG. CAGD&CG. Input ... – PowerPoint PPT presentation

Number of Views:491
Avg rating:3.0/5.0
Slides: 91
Provided by: Tur48
Category:
Tags: clouds | curve | fitting | point

less

Transcript and Presenter's Notes

Title: Curve fitting to point clouds


1
Curve fitting to point clouds
  • Reporter Lincong Fang
  • Oct 18, 2006

2
Curve fitting
  • The data points are ordered.

3
Curve fitting to point clouds
  • The data points are unorganized.

4
Applications
  • Some applications
  • Reverse engineering
  • Curve design
  • Surface reconstruction
  • Etc.

5
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

6
Curve Reconstruction from unorganized points
  • In-Kwon Lee
  • CAGD 2000

7
Least squares
8
Moving Least Squares
9
Moving Least Squares
10
Moving Least Squares
11
Moving Least Squares
12
The choice of H
13
Improved Moving Least Squares
Delaunay triangulation
Euclidean minimum spanning tree (EMSP)
14
(No Transcript)
15
Correlation
16
Correlation
17
(No Transcript)
18
(No Transcript)
19
Refining
20
Compare with and without EMST
21
Ordering points
22
Example
23
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

24
Curve reconstruction based on interval B-spline
curve
  • Hongwei Lin, Wei Chen,
  • Guojin Wang
  • The Visual Computer,21(6), 418-427,2005

25
Overview
26
(No Transcript)
27
(No Transcript)
28
Shape-based joining scheme
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
Sequence Joining Method
34
Boundary sequence
35
Example
36
Example
37
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

38
Grouping and parameterizing irregularly spaced
points for curve fitting
  • Ardeshir Goshtasby
  • ACM Transactions on Graphics, 19185--203, 2000

39
(No Transcript)
40
Minor and major ridges
41
Map into a digital image
42
Minor and major ridges
43
Example
44
Example
45
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

46
Multidimensional curve fitting to unorganized
data points by nonlinear minimization
  • Lian Fang, David C Gossard
  • CAD 95

47
(No Transcript)
48
Physical analogy
49
(No Transcript)
50
Error term
51
Example
52
Fitting B-spline curves to point clouds by
curvature-based squared distance minimization
  • Wenping Wang, Helmut Pottmann, Yang Liu
  • ToG 2006

53
(No Transcript)
54
(No Transcript)
55
Point distance minimization
56
Tangent distance minimization
57
Squared distance minimization
Pottman 2003
58
Squared distance minimization
59
Squared distance minimization
60
Comparison
  • PDM has slow convergence
  • TDM has fast but unstable convergence
  • SDM yields a more balanced performance between
    efficiency and stability

61
Comparison
The fitting curve generated by PDM, TDM, SDM in
50 iterations
Initial curve
62
Foot point computation
63
Open curves
64
Initial curves and control points
  • Specify by user
  • Compute a quadtree partition of
  • the data points
  • Automatic or specify by user, and adjustment
    (Yang 2004)

65
Example
66
Reconstructing B-spline curves from point
cloudsA tangential flow approach using least
squares minimization
  • Yang Liu, Huaiping Yang, Wenping Wang
  • Shape Modeling and Applications, 2005
    International Conference

67
(No Transcript)
68
Input
  • Unacceptable point clouds.

69
Data Analysis
70
Initialization and approximation
Random point S
B-spline curve
Fitting line L
I
71
Growing
72
Knot insertion
  • All points are handled, add a knot where the
    maximum error occurs
  • Else insert a knot and redistribute all the knots
    and make them equally spaced

73
Finding projection points
Sharp corners
74
Filtering points
T
75
Other cases
Less control points
EMST with wrong topology
Very sharp corner
76
Example
77
Example
78
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

79
Fitting unorganized point clouds with active
implicit B-spline curves
  • Zhouwang Yang, Jiansong Deng, Falai Chen
  • Visual Computer 2005

80
(No Transcript)
81
(No Transcript)
82
(No Transcript)
83
(No Transcript)
84
Example
85
(No Transcript)
86
(No Transcript)
87
Approaches overview
  • Preprocess the point clouds
  • Thin the point clouds (Levin98, Lee00)
  • Point clusters (Lin 04)
  • Map into a digital image (Goshtasby00)
  • Mathematical model
  • Parameteric curves
  • Implicit curves
  • Other methods

88
Conclusion
  • Complex topology
  • Digital image
  • Implicit curves
  • Tangential flow
  • Initial curves
  • Parameteric curves
  • Implicit curves

89
Problems and future work
  • Knot insertion
  • Foot point compute
  • Singular points
  • Surface reconstruction

90
Thank you!
Write a Comment
User Comments (0)
About PowerShow.com