Lec 4 Scalar Field Visualization

Scalar Field Visualization
Leif Kobbelt
Types of Data
dimension of nD
data type
3D
2D
1D
0D
1D
2
Visual Computing Institute | Prof. Dr. Leif Kobbelt

Computer Graphics and Multimedia
Data Analysis and Visualization
2D
3D
nD
dimension of
domain
Function Plots and Height Fields
Visualization of 1D or 2D scalar fields
1D scalar field:
IR IR
2D scalar field:
IR 2 IR

Function plot for a 1D scalar field

2D Points
{( s, f ( s )) | s IR }
1D manifold: line

Function plot for a 2D scalar field

3D Points
{( s, t , f ( s, t ) | ( s, t ) IR 2 }
2D manifold: surface
Surface representations
Wireframe
Hidden lines
Shaded surface

The Geometry of Images



Isolines
Visualization of 2D scalar fields
Given a scalar function f : IR
and a scalar value
Isoline consists of points
{( x, y ) | f ( x, y ) c}
If f() is differentiable and grad(f) 0, then isolines are

curves
Contour lines

Isolines
10

10
Isolines
11

11
Isolines
pixel by pixel contouring
straightforward approach: scanning all pixels for isovalue
input
f : (1,...,xmax) x (1,...,ymax) R
Isovalues I1,..., In and isocolors c1,...,cn
algorithm
for all (x,y) (1,...,xmax) x (1,...,ymax) do
for all k { 1,...,n } do
if |f(x,y)-Ik| < then
draw(x,y,ck)
problem: isolinie can be missed if the gradient of f() is too large
(despite range )
12

12
Isolines
Marching squares
13
representation of the scalar function on a rectilinear grid

scalar values are given at each vertex ffij
take into account the interpolation within cells
isolines cannot be missed
divide and conquer: consider cells independently of each other

13
Isolines
Which cells will be intersected ?
Initially mark all vertices by + or , depending on the

conditions fij c , fij < c
No isoline passes through cells (=rectangles) which have

the same sign at all four vertices
So we only have to determine the edges with different signs
14

14
Isolines
Only 4 different cases (classes) of combinations of signs
Symmetries: rotation, reflection, change + Compute intersections between isoline and cell edge,
based on linear interpolation along the cell edges
+
?
-
15

15
Isolines
We can distinguish the cases by a decider
Mid point decider
Interpolate the function value in the center
f center
1
( f i , j f i 1, j f i , j 1 f i 1, j 1 )
4
If fcenter < c we chose the right case, otherwise we chose the left case
+
-
Not always correct solution
16

16
Isolines
Asymptotic decider
Consider the bilinear interpolant within a cell

The true isolines within a cell are hyperbolas
Cell boundaries
Asymptote
Asymptote
17

17
Isolines
Interpolate the function bilinearly
f ( x, y ) f i , j (1 x)(1 y ) f i 1, j x(1 y ) f i , j 1 (1 x) y f i 1 j 1 xy
Transform f() to
f ( x, y ) ( x x0 )( y y0 )
is the function value in the

intersection point
of the asymptotes
18

18
(x0, y0)
Isolines
If c we chose the right case, otherwise we

chose the left one
+
+
-
19

19
Isolines
Explicit transformation f() to
f ( x, y ) ( x x0 )( y y0 )
can be avoided
Idea: investigate the order of intersection points either
along x or y axis
Build pairs of first two and last two intersections
not possible
valid
20

20
3D Scalar Field Visualization
21

21
Volumetric Representations
Implicit representation F(x,y,z) = 0

Signed distance function
Sample on a uniform cartesian grid
Trilinear interpolation
22
22

F011
F001
F111
F101
F110
F000
23
23
F100

F(i+u,j+v,k+w) =
F000 (1-u)(1-v)(1-w) +
F100 u (1-v)(1-w) +
F010 (1-u) v (1-w) +
...
F111
u v w
Indirect Rendering:
Isosurface extraction: Marching Cubes
Direct Rendering:
Ray Casting
24
24

Volume Data
discrete representation:
Fijk = F(i x, j y, k z)
piecewise polynomial (linear) interpolation

tri-linear functions per cell (voxel)
algebraic iso-surfaces of degree 3
approximation by a polygonal mesh
25
25

Surface Extraction
Find point samples on the iso-surface

exploit voxel structure
use function values
Connect neighboring samples
exploit voxel neighborhood structure
relation between voxels and polygons
26
26

Surface Sampling
in which voxel ?
27
27

Surface Sampling
in which voxel ?
F((i+u) x, (j+v) y, (k+w) z) =
(1-u)(1-v)(1-w)Fijk + (1-u)(1-v) w Fijk+1

+
(1-u) v (1-w)Fij+1k + (1-u) v w Fij+1k+1 +
u (1-v)(1-w)Fi+1jk + u (1-v) w Fi+1jk+1 +
u
v (1-w)Fi+1j+1k+ u v w Fi+1j+1k+1 +
(u,v,w) [0,1]3 min Fijk F(u,v,w) max Fijk
28
28

Surface Sampling
in which voxel ?
Sc[F] passes through all (and only) voxels
with min Fijk c max Fijk
29
29

Surface Sampling
in which voxel ?
how to find these voxels efficiently ?
30
30

Surface Sampling
in which voxel ?
how to find these voxels efficiently ?
- store min/max {Fijk,Fijk+1,Fij+1k,Fij+1k+1,
Fi+1jk,Fi+1jk+1,Fi+1j+1k,Fi+1j+1k+1}
- build octree
31
31

Min/Max Tree
store min/max for every voxel Vijkn

octree Vijkn-1 has children
V2i 2j 2kn ... V 2i+1 2j+1 2k+1n
propagate min/max values upwards
surface extraction:
traverse only the relevant voxels
O(n3) O(n2 log n)
32
32

Surface Sampling
per voxel : ray intersection

ray ... P + Q
intersection ... G() = F(P + Q) = c

solve cubic equation ...
33
33

Surface Sampling
per voxel : ray intersection

ray ... P + Q
intersection ... G() = F(P + Q) = 0

solve cubic equation ...
ray parallel to coordinate axis
solve linear equation ...
34
34

Problems / Bad Approximation
35
35

Surface Sampling
use additional information to improve

iso-surface approximation
normal vector to iso-surface Sc[F]:
F(x,y,z) = [ F/x, F/y, F/z ]
find surface samples by (approximately)
intersecting tangent planes
36
36

Error Quadrics
squared distance to plane:
T
T
v x, y, z ,1 ,
p a , b, c , d
2
2
T
dist p, v p , v vT pp T v : vT QP v
a 2 ab ac ad
2
ab b bc bd
QP
ac bc c 2 cd
2
ad bd cd d
sum distances to vertex' planes:
dist p, v
2
planes p
37
37
planes p

vT Q p v vT
T
Q
v
:
v
Qv v
p
planes p
Error Quadrics
optimal point that minimizes the error:
T
n p n p v d p n p
p a p , bp , c p , d p
:
n
with
n
1
p
p
solve using pseudo-inverse (SVD)

(underdetermined vs. overdetermined)
38
38

Surface Data
polygon meshes / triangle meshes

M = ({pi},{Tj})
compute samples
on grid edges
within voxels
compute faces (vertex connectivity)
for each voxel
for each grid edge
39
39

Marching Cubes [Lorensen/Cline 87]
extract surface patch for every grid cell

classification by the signs at the corners
(black, white, gray)
approximate samples
linear interpolation of scalar distance values
along cell edges
look-up table with pre-computed triangulations
28 entries, many symmetries
40
40

Visit Each Gray Cell ...
41
41

Compute Samples on Edges
d i , j ,k
i
, j, k
d
i
,
j
,
k
i
1
,
j
,
k
42
42

d i , j ,k
i, j
,k
d
i
,
j
,
k
i
,
j
1
,
k
43
43

d i , j ,k
i, j , k
d i , j ,k d i , j ,k 1
44
44

Lookup Mesh Connectivity
...
45
45

Cell Connectivity Table
46
46

Extended Marching Cubes
same algorithmic principle

for each cell ...
Find surface samples on the grid cell edges
Evaluate surface normals at the sample points
Feature detection
Feature sampling
Feature reconstruction
47
47

Feature Classification
Normal cone
for a smooth (flat)
non-feature region
48
48

Normal cone
for a smooth (flat)
non-feature region
49
49

Normal cone
for an edge feature
(one large angle)
50
50

51
51

Normal cone
for a corner feature
(two large angles)
52
52

Feature Detection
opening angles of the normal cone

= maxi,j <ni, nj>
= mink <nk, ni x nj>
edge features
> 0, > 0 sharp edge
corner features
> 0, <= 0 sharp corner
53
53

Feature Sampling
Each pair ( pi , ni) defines a tangent element

Intersect tangent planes to approximate
a piecewise smooth surface
Find least squares solution by SVD
Overdetermined cases
Underdetermined cases
Suppress smallest singular value at edge
features
54
54

Lookup Mesh Connectivity
...
55
55

Modified Lookup Table
Generate triangle fans centered around the

feature sample
...
56
56

Feature Reconstruction
feature
detection
57
57
feature
sampling

edge-flipping
Results
Mesh complexity
Two additional triangles per feature sample
Typically 10% overhead
Computation time
20% to 40% overhead
Improved approximation
Parameters 0 and 0
False positive feature detection !?
Applications ...
58
58

CSG ... CAD / CAM
65 x 65 x 65
59
59

CSG ... Milling Simulation
257 x 257 x 257
60
60

65 x 65 x 65
Iso-Surface Extraction
61
61

Iso-Surface Extraction
62
62

Indirect Rendering:
Isosurface extraction: Marching Cubes
Direct Rendering:
Ray Casting
63
63

Volume Rendering
64
64

Volume Rendering
65
65

Volume Rendering
66
66

Volume Rendering
67
67

Ray Casting
Maximum Intensity Projection
Maximum Smax
Scalar value S
68
68

ray
Maximum Intensity Projection
69
69

Absorption
Ray
I(s0)
s0
70
70

Absorption
Ray
I(s0)
s
s0
I(s0)
s0
71
71
Ray
q(s)

Ray Casting
Image order: for each pixel ...

Shoot rays from eye through each image plane
pixel
Integrate color c & opacity along the ray r

s
c(r ) c( s ) e
0
72
72

( t ) dt
0
ds
73
73

Opacity
base color c, opacity
damping factor 1- = e-
emitted color c
resulting (effective) color
c + (1-) cbackground
74
74

Ray Casting
set color c(.) and opacity (.) for all voxels

for all pixels ui = (ui, vi)
find ray from eye through pixel ui : eye + d
init cacc(ui) = acc(ui) = 0
for all samples xj = (xj, yj, zj) = eye + j d
get c(xj) and (xj) by tri-linear interpolation
composite with cacc(ui) and acc(ui)
75
75

Opacity Accumulation
Continous compositing
L
c(ui ) c(E D) e
( E tD ) dt
0
Discrete compositing
c(ui ) c( x j ) ( x j ) (1 ( xm ))
j 0
m 0
with c( xk ) cbkg , ( xk ) 1
K
76
76

j 1
cK

K
77
77
ci

i

c0

0
C[0..K]
cK

K
78
78
ci

i

c0

0
C[1..K]
cK

K
79
79
ci

i

back-to-front
c0

0
C[0..K-1]
cK

K
80
80
ci

i

front-to-back
c0

0
Back-to-Front Compositing
j 1
c( x j ) ( x j ) (1 ( xm ))
j 0
m 0
c( x0 ) ( x0 )
j 1
c( x j ) ( x j ) (1 ( xm )) (1 ( x0 ))
j 1
m 1
81
81

Front-to-Back Compositing
Iterative accumulation j = 0 .. K
C C c j j (1 )
j (1 )
with
j 1
(1 ) (1 m )
m 0
82
82

Transfer Functions
how to define the

base color c
opacity
for each voxel ?
CT, MRI, ...

diffusion tensor imaging
83
83

Transfer Functions
scalar value s
T(s)
emission RGB
absorption A
84
84

Transfer Functions
85
85

Transfer Functions
86
86

Transfer Functions
87
87

Transfer Functions
+ Shading
88
88

Transfer Functions
+ Shading
+ Color coding
89
89

Absorption vs. MIP
90
90

Intensity
Compositing Schemes
Max
Accumulate
Average
First
Depth
91

91
Compositing Schemes
Intensity
Compositing: First
Extracts isosurfaces
First
Depth
92

92
Compositing Schemes
Intensity
Compositing: Average
Produces basically an X-ray picture
Average
Depth
93

93
Compositing Schemes
Intensity
Maximum Intensity Projection (MIP)

Often used for magnetic resonance angiograms
Good to extract vessel structures
Max
Depth
94

94
Compositing Schemes
Intensity
Compositing: Accumulate
Emission-absorption model
Make transparent layers visible (- classification)
Accumulate
Depth
95

95
Volume Shading
Lighting / Shading
Phong model, Gouraud shading
F ( x , y , z )
Estimate normal by N ( x, y, z )
F ( x , y , z )
Approximate F(x,y,z) by central difference
F ( xi 1 , yi , zi ) F ( xi 1 , yi , zi )
1
F ( x, y, z ) F ( xi , yi 1 , zi ) F ( xi , yi 1 , zi )
2
F
(
x
,
y
,
z
)
F
(
x
,
y
,
z
)
i 1
i
i
i
i
i 1
96
96

Volume Rendering Modes
Classification
assign spatial opacity values (xi, yi, zi)
assign for isovalues F(x,y,z) = (transfer
function)
isosurface ... = (=0) ... blur with radius r
1
,
F ( xi ) 0 F ( xi ) v
1 F ( xi ) v
( xi ) v 1
, F ( xi ) 0 F ( xi ) v F ( xi )
r F ( xi ) ,
otherwise
0
97
97

98
98

rF
99
99

rF
100
100

Ray Casting
Acceleration techniques
Spatial data structures (octree)
Find and skip empty regions
Find homogenous regions, use lower sample
rate
Early ray termination
Opacity threshold terminates ray traversal
Requires front-to-back traversal
Fast cell traversal (Bresenham-like)
101
101


Lec 4 Scalar Field Visualization

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Lec 4 Scalar Field Visualization

Caricato da

Copyright:

Formati disponibili

Scalar Field Visualization

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Function Plots and Height Fields

Visualization of 1D or 2D scalar fields

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Function Plots and Height Fields

Function plot for a 1D scalar field

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Function Plots and Height Fields

Function plot for a 2D scalar field

Visual Computing Institute | Prof. Dr. Leif Kobbelt

The Geometry of Images

Visual Computing Institute | Prof. Dr. Leif Kobbelt

The Geometry of Images

Visual Computing Institute | Prof. Dr. Leif Kobbelt

The Geometry of Images

Visual Computing Institute | Prof. Dr. Leif Kobbelt

If f() is differentiable and grad(f) 0, then isolines are

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

representation of the scalar function on a rectilinear grid

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Initially mark all vertices by + or , depending on the

No isoline passes through cells (=rectangles) which have

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Interpolate the function value in the center

Not always correct solution

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Consider the bilinear interpolant within a cell

Visual Computing Institute | Prof. Dr. Leif Kobbelt

is the function value in the

Visual Computing Institute | Prof. Dr. Leif Kobbelt

If c we chose the right case, otherwise we

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

3D Scalar Field Visualization

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Implicit representation F(x,y,z) = 0

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

piecewise polynomial (linear) interpolation

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Find point samples on the iso-surface

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

(1-u)(1-v)(1-w)Fijk + (1-u)(1-v) w Fijk+1

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Visual Computing Institute | Prof. Dr. Leif Kobbelt

store min/max for every voxel Vijkn

Visual Computing Institute | Prof. Dr. Leif Kobbelt

per voxel : ray intersection

intersection ... G() = F(P + Q) = c

Visual Computing Institute | Prof. Dr. Leif Kobbelt

per voxel : ray intersection

intersection ... G() = F(P + Q) = 0

Visual Computing Institute | Prof. Dr. Leif Kobbelt

Problems / Bad Approximation

Visual Computing Institute | Prof. Dr. Leif Kobbelt