Gary%20Miller - PowerPoint PPT Presentation

About This Presentation
Title:

Gary%20Miller

Description:

Size Competitive Meshing without Large Angles. Gary L. Miller. Carnegie Mellon ... bounded largest angles. Can be efficiently computed by Delaunay Refinement ... – PowerPoint PPT presentation

Number of Views:57
Avg rating:3.0/5.0
Slides: 108
Provided by: Don181
Learn more at: http://www.cs.cmu.edu
Category:
Tags: 20miller | angles | gary

less

Transcript and Presenter's Notes

Title: Gary%20Miller


1
Size Competitive Meshing without Large Angles
Gary L. Miller Carnegie Mellon Computer
Science Joint work with Todd Phillips and Don
Sheehy
2
The Problem
Input A Planar Straight Line Graph
3
The Problem
Input A Planar Straight Line Graph
Output A Conforming Triangulation
4
The Problem
Input A Planar Straight Line Graph
Output A Conforming Triangulation
5
The Problem
Input A Planar Straight Line Graph
Output A Conforming Triangulation
Quality
6
What is a quality triangle?
7
What is a quality triangle?
No Large Angles
No Small Angles
8
What is a quality triangle?
No Large Angles
No Small Angles
Implies triangles have bounded aspect ratio.
9
What is a quality triangle?
No Large Angles
No Small Angles
Implies triangles have bounded aspect ratio.
Implies triangles have bounded largest angles.
10
What is a quality triangle?
No Large Angles
No Small Angles
Implies triangles have bounded aspect ratio.
Implies triangles have bounded largest angles.
Can be efficiently computed by Delaunay Refinement
11
What is a quality triangle?
No Large Angles
No Small Angles
Sufficient for many applications.
Implies triangles have bounded aspect ratio.
Implies triangles have bounded largest angles.
Can be efficiently computed by Delaunay Refinement
12
What is a quality triangle?
No Large Angles
No Small Angles
Sufficient for many applications.
Implies triangles have bounded aspect ratio.
Can be asymptotically smaller then Delaunay
Refinement triangulations.
Implies triangles have bounded largest angles.
Can be efficiently computed by Delaunay Refinement
13
What is a quality triangle?
No Large Angles
No Small Angles
Sufficient for many applications.
Implies triangles have bounded aspect ratio.
Can be asymptotically smaller then Delaunay
Refinement triangulations.
Implies triangles have bounded largest angles.
More difficult to analyze.
Can be efficiently computed by Delaunay Refinement
14
In Defense of Quality
15
In Defense of Quality
16
In Defense of Quality
17
In Defense of Quality
What went wrong?
18
In Defense of Quality
What went wrong?
19
Interpolation Problem
No Large Angles
Large Angles
Large angles give large H1 errors that FEMs try
to minimize.
20
Paying for the spread
21
Paying for the spread
Spread L/s
L
s
22
Paying for the spread
Optimal No-Large-Angle Triangulation
23
Paying for the spread
What if we dont allow small angles?
24
Paying for the spread
What if we dont allow small angles?
25
Paying for the spread
What if we dont allow small angles?
O(L/s) triangles!
26
Paying for the spread
What if we dont allow small angles?
O(L/s) triangles!
27
Delaunay Refinement
28
Delaunay Refinement
29
Delaunay Refinement
  • Theorem Delaunay Refinement on point sets
    terminates and returns a triangulation with
  • all angles at least 30-e degrees
  • O(n log L/s) triangles

30
Patersons Example
  • Requires O(n2) points.

O(n) points
O(n) lines
31
Patersons Example
  • Requires ?(n2) points.

O(n) points
O(n) lines
32
Patersons Example
  • Requires ?(n2) points.

O(n) points
O(n) lines
33
Patersons Example
  • Requires ?(n2) points.

O(n) points
O(n) lines
34
Patersons Example
  • Requires ?(n2) points.

O(n) points
O(n) lines
35
Patersons Example
  • Requires ?(n2) points.

O(n) points
O(n) lines
36
Past Results
  • O(n) triangles with 90o largest angles for
    polygons with holes. Bern, Mitchell, Ruppert,
    95
  • ?(n2) lower bound for arbitrary PLSGs.
    Paterson
  • O(n2) triangles with 132o angles on PSLGs. Tan,
    96

37
Past Results
Delaunay Refinement Methods
No-Large-Angle Methods
Cons
Pros
Cons
Pros
Smaller Meshes Not well-graded Complicated to
Implement
Good Theory Optimal Runtime Graded Mesh Simple to
Implement Esthetically Nice
Huge Meshes O(L/s) Require Hacks to handle small
input angles. Size depends on smallest angle.
Smaller Meshes Worst-Case Optimal

38
Past Results
OUR
Delaunay Refinement Methods
No-Large-Angle Methods
Cons
Pros
Cons
Pros
Smaller Meshes Not well-graded Complicated to
Implement
Good Theory Optimal Runtime Graded Mesh Simple to
Implement Esthetically Nice
Huge Meshes O(L/s) Require Hacks to handle small
input angles. Size depends on smallest angle.
Smaller Meshes Worst-Case Optimal
39
Past Results
OUR
Delaunay Refinement Methods
No-Large-Angle Methods
Cons
Pros
Cons
Pros
Only Worst-Case Bounds Not well-graded Complicated
to Implement
Good Theory Optimal Runtime Graded Mesh Simple to
Implement Esthetically Nice
Huge Meshes O(L/s) Require Hacks to handle small
input angles. Size depends on smallest angle.
Smaller Meshes Worst-Case Optimal
40
Past Results
OUR
Delaunay Refinement Methods
No-Large-Angle Methods
Cons
Pros
Cons
Pros
Only Worst-Case Bounds Not well-graded Complicated
to Implement
Good Theory Optimal Runtime Graded Mesh Simple to
Implement Esthetically Nice
Huge Meshes O(L/s) Require Hacks to handle small
input angles. Size depends on smallest angle.
Smaller Meshes Worst-Case Optimal size
Log L/s -competitive
Graded on Average
41
Past Results
OUR
Delaunay Refinement Methods
No-Large-Angle Methods
Cons
Pros
Cons
Pros
Only Worst-Case Bounds Not well-graded Complicated
to Implement
Good Theory Optimal Runtime Graded Mesh Simple to
Implement Esthetically Nice
Huge Meshes O(L/s) Require Hacks to handle small
input angles. Size depends on smallest angle.
Smaller Meshes Worst-Case Optimal size
Log L/s -competitive
Graded on Average
Our Angle bounds are not as good, 170o versus
140o
42
Local Feature Size
lfs(x) distance to second nearest vertex.
Note lfs is defined on the whole plane.
x
lfs(x)
43
The OSM Algorithm(Overlay Stitch Meshing)
44
The OSM Algorithm(Overlay Stitch Meshing)
45
The OSM Algorithm(Overlay Stitch Meshing)
46
The OSM Algorithm(Overlay Stitch Meshing)
47
The OSM Algorithm(Overlay Stitch Meshing)
48
The OSM Algorithm(Overlay Stitch Meshing)
49
The OSM Algorithm(Overlay Stitch Meshing)
50
The OSM Algorithm(Overlay Stitch Meshing)
51
The OSM Algorithm(Overlay Stitch Meshing)
52
The OSM Algorithm(Overlay Stitch Meshing)
53
The OSM Algorithm(Overlay Stitch Meshing)
54
The OSM Algorithm(Overlay Stitch Meshing)
55
The OSM Algorithm(Overlay Stitch Meshing)
56
The OSM Algorithm(Overlay Stitch Meshing)
  • An Overlay Edge is kept if
  • 1. It does not intersect the input,
  • OR

57
The OSM Algorithm(Overlay Stitch Meshing)
  • An Overlay Edge is kept if
  • 1. It does not intersect the input,
  • OR
  • 2. It forms any good intersection with the input.

58
The OSM Algorithm(Overlay Stitch Meshing)
  • An Overlay Edge is kept if
  • 1. It does not intersect the input,
  • OR
  • 2. It forms any good intersection with the input.

at least 30o
59
Angle Guarantees
60
Angle Guarantees
We want to prove that the stitch vertices that we
keep do not form bad angles.
61
Angle Guarantees
An Overlay Triangle
62
Angle Guarantees
An Overlay Triangle
An Overlay Edge
63
Angle Guarantees
Gap Ball
An Overlay Triangle
An Overlay Edge
64
Angle Guarantees
A Good Intersection
65
Angle Guarantees
A Bad Intersection
66
Angle Guarantees
A Bad Intersection
67
Angle Guarantees
A Bad Intersection
68
Angle Guarantees
How Bad can it be?
A Bad Intersection
69
Angle Guarantees
How Bad can it be?
70
Angle Guarantees
How Bad can it be?
71
Angle Guarantees
How Bad can it be?
About 10o
72
Angle Guarantees
How Bad can it be?
Theorem If any input edge makes a good
intersection with an overlay edge then any other
intersection on that edge is not too bad.
About 10o
73
Size Bounds
How big is the resulting triangulation?
74
Size Bounds
How big is the resulting triangulation?
Goal log(L/s)-competitive with optimal
75
Size Bounds
How big is the resulting triangulation?
Goal log(L/s)-competitive with optimal
Overlay Phase O(n log L/s) points added
76
Size Bounds
How big is the resulting triangulation?
Goal log(L/s)-competitive with optimal
Overlay Phase O(n log L/s) points
Stitching Phase O(sE lfs0-1(z)dz)
77
Rupperts Idea
  • r Q(lfs(c))
  • area(D) Q(lfs(c)2)
  • of triangles ?(ss lfs(x,y)-2 dxdy)

r
c
78
Rupperts Idea
  • r Q(lfs(c))
  • area(D) Q(lfs(c)2)
  • of triangles ?(ss lfs(x,y)-2 dxdy)

r
c
Caveat Only Works for well-graded meshes with
bounded smallest angle.
79
Extending Rupperts Idea
of triangles ?(ss lfs(x,y)-2 dxdy)
80
Extending Rupperts Idea
of triangles ?(ss lfs(x,y)-2 dxdy)
An input edge e intersecting the overlay mesh
81
Extending Rupperts Idea
of triangles ?(ss lfs(x,y)-2 dxdy)
of triangles along e ?(sz2 e lfs(z)-1 dz)
An input edge e intersecting the overlay mesh
82
Extending Rupperts Idea
of triangles ?(ss lfs(x,y)-2 dxdy)
of triangles along e ?(sz2 e lfs(z)-1 dz)
Now we can compute the size of our output mesh.
We just need to compare these integrals with
optimal.
83
Competitive Analysis
Warm-up
84
Competitive Analysis
Warm-up
90o
85
Competitive Analysis
Warm-up
?o
86
Competitive Analysis
Warm-up
87
Competitive Analysis
Warm-up
88
Competitive Analysis
Warm-up
89
Competitive Analysis
Warm-up
90
Competitive Analysis
Warm-up
91
Competitive Analysis
Warm-up
Suppose we have a triangulation with all angles
at least 170o.
92
Competitive Analysis
Warm-up
Suppose we have a triangulation with all angles
at least 170o.
an optimal
93
Competitive Analysis
Warm-up
Suppose we have a triangulation with all angles
at least 170o.
an optimal
Every input edge is covered by empty lenses.
94
Competitive Analysis
This is what we will integrate over to get a
lower bound.
Suppose we have a triangulation with all angles
at least 170o.
the optimal
Every input edge is covered by empty lenses.
95
Competitive Analysis
This is what we will integrate over to get a
lower bound.
We show that in each lens, we put O(log (L/s))
points on the edge.
96
Competitive Analysis
This is what we will integrate over to get a
lower bound.
e
We show that in each lens, we put O(log (L/s))
points on the edge.
97
Conclusion
  • A new algorithm for no-large-angle triangulation.
  • The output has bounded degree triangles.
  • The first log-competitive analysis for such an
    algorithm.

98
Conclusion
  • A new algorithm for no-large-angle triangulation.
  • The output has bounded degree triangles.
  • The first log-competitive analysis for such an
    algorithm.
  • We used simple calculus to bound mesh sizes.

99
Where to go from here?
  • 3D?

100
Where to go from here?
  • 3D?
  • Better angle bounds?

101
Where to go from here?
  • 3D?
  • Better angle bounds?
  • Find a constant competitive algorithm.

102
Where to go from here?
  • 3D?
  • Better angle bounds?
  • Find a constant competitive algorithm.

Thanks
103
Competitive Analysis
sz2 e 1/lfs(z) dz O(log (L/s))
e
104
Competitive Analysis
sz2 e 1/lfs(z) dz O(log (L/s))
e
Parametrize e as 0, l where l length(e)
105
Competitive Analysis
sz2 e 1/lfs(z) dz O(log (L/s))
e
Parametrize e as 0, l
1st trick lfs \ge s everywhere
106
Competitive Analysis
sz2 e 1/lfs(z) dz O(log (L/s))
t
e
Parametrize e as e(t) for t2 0, l l
length(e)
1st trick lfs s everywhere
2nd trick lfs ct for t2 0,l/2
107
Competitive Analysis
sz2 e 1/lfs(z) dz O(log (L/s))
t
e
Parametrize e as e(t)
? z2 e 1/lfs(z) dz 2 ?s 0 1/s 2 ? sl/2 1/x
dx O(1) O(log(l/2) log s) O(log l/s)
O(log L/s)
1st trick lfs s everywhere
2nd trick lfs ct for t2 0,l/2
Write a Comment
User Comments (0)
About PowerShow.com