geometric active contours ron kimmel ron computer science department technion-israel institute of...
TRANSCRIPT
![Page 1: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/1.jpg)
Geometric Active ContoursRon Kimmel
www.cs.technion.ac.il/~ron
Computer Science Department Technion-Israel Institute of Technology
Geometric Image Processing Lab
![Page 2: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/2.jpg)
Edge Detection Edge Detection:
The process of labeling the locations in the image where the gray level’s “rate of change” is high. OUTPUT: “edgels” locations,
direction, strength
Edge Integration: The process of combining “local” and perhaps sparse and
non-contiguous “edgel”-data into meaningful, long edge curves (or closed contours) for segmentation OUTPUT: edges/curves consistent with the local data
![Page 3: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/3.jpg)
The Classics Edge detection:
Sobel, Prewitt, Other gradient estimators Marr Hildreth
zero crossings of Haralick/Canny/Deriche et al.
“optimal” directional local max of derivative
Edge Integration: tensor voting (Rom, Medioni, Williams, …) dynamic programming (Shashua & Ullman) generalized “grouping” processes (Lindenbaum et al.)
IG *
![Page 4: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/4.jpg)
The “New-Wave” Snakes Geodesic Active Contours Model Driven Edge Detection
Edge Curves
“nice” curves that optimize a functional of g( ), i.e.
nice: “regularized”,
smooth, fit some prior information
curve
dsg ) (
Image
),( yxg
Edge Indicator Function
2
1
1 | ( * ) |G I
![Page 5: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/5.jpg)
Geodesic Active Contours
Snakes Terzopoulos-Witkin-Kass 88 Linear functional efficient implementation non-geometric depends on parameterization
Open geometric scaling invariant, Fua-Leclerc 90 Non-variational geometric flow
Caselles et al. 93, Malladi et al. 93 Geometric, yet does not minimize any functional
Geodesic active contours Caselles-Kimmel-Sapiro 95 derived from geometric functional non-linear inefficient implementations:
Explicit Euler schemes limit numerical step for stability Level set method Ohta-Jansow-Karasaki 82, Osher-Sethian 88
automatically handles contour topology Fast geodesic active contours Goldenberg-Kimmel-Rivlin-
Rudzsky 99 no limitation on the time step efficient computations in a narrow band
![Page 6: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/6.jpg)
Laplacian Active Contours Closed contours on vector fields
Non-variational models Xu-Prince 98, Paragios et al. 01 A variational model Vasilevskiy-Siddiqi 01
Laplacian active contours open/closed/robust Kimmel-Bruckstein 01
Most recent:variational measures
for good old operatorsKimmel-Bruckstein 03
![Page 7: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/7.jpg)
Segmentation
![Page 8: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/8.jpg)
Segmentation
Ultrasound images
Caselles,Kimmel, Sapiro ICCV’95
![Page 9: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/9.jpg)
Segmentation
Pintos
![Page 10: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/10.jpg)
Woodland Encounter Bev Doolittle 1985
With a good prior who needs the data…
![Page 11: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/11.jpg)
Segmentation
Caselles,Kimmel, Sapiro ICCV’95
![Page 12: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/12.jpg)
Prior knowledge…
![Page 13: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/13.jpg)
Prior knowledge…
![Page 14: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/14.jpg)
Segmentation
![Page 15: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/15.jpg)
Segmentation
![Page 16: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/16.jpg)
Segmentation
Caselles,Kimmel, Sapiro ICCV’95
![Page 17: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/17.jpg)
Segmentation
With a good prior who needs the data…
![Page 18: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/18.jpg)
Wrong Prior???
![Page 19: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/19.jpg)
Wrong Prior???
![Page 20: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/20.jpg)
Wrong Prior???
![Page 21: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/21.jpg)
Curves in the Plane C(p)={x(p),y(p)}, p [0,1]
y
x
C(0)
C(0.1) C(0.2)
C(0.4)
C(0.7)
C(0.95)
C(0.9)
C(0.8)
pC =tangent
![Page 22: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/22.jpg)
Arc-length and Curvature
s(p)= | |dp 0
p
| | 1,sC pC | |p
sp
CC
C
ssC N
1
ssC N
C
![Page 23: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/23.jpg)
Calculus of Variations
Find C for which is an extremum
Euler-Lagrange:
1
0
( , )pE L C C dp( ) ( , , , )p p pE L x y x y dp
0
( )px x x
E dE L L dp
dp
0
0
p
p
x x
y y
dL L
dp
dL L
dp
![Page 24: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/24.jpg)
Calculus of Variations
Important Example Euler-Lagrange: , setting Curvature flow
( )t ss ssC C C N
1
0
| |pE C dp| | 0pN C
QQQQQQQQQQQQQQ| |pds C dp
0N
![Page 25: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/25.jpg)
Potential Functions (g)
x
I(x,y) I(x)
x
g(x)
xx
g(x,y)
Image
Edges2
1
1 | ( * ) |G I
![Page 26: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/26.jpg)
Snakes & Geodesic Active Contours
Snake modelTerzopoulos-Witkin-Kass 88
Euler Lagrange as a gradient descent
Geodesic active contour modelCaselles-Kimmel-Sapiro 95
Euler Lagrange gradient descent
CgCCdt
dCpppppp
0arg min
L C
Cg C ds
NΝg(C),κCgdt
dC
1
0
22 )(2minarg dpCgCC ppp
C
![Page 27: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/27.jpg)
Maupertuis Principle of Least Action
Snake = Geodesic active contourup to some , i.e Snakes depend on parameterization. Different initial parameterizations yield solutions for different geometric functionals
0g g E 0E
12
0
0
0
arg min ( ( ( )) | | )
arg min ( ( ))
pC
L
C
g C p C dp
g C s E ds
x
y
p
1
0
Caselles Kimmel Sapiro, IJCV 97
![Page 28: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/28.jpg)
Geodesic Active Contours in 1D
Geodesic active contours are reparameterization invariant
I(x)
x
g(x)
x
NΝ
),(CgCgdt
dC
![Page 29: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/29.jpg)
Geodesic Active Contours in 2D
2
1
1 | ( * ) |G I g(x)=
G *I
NNCgCgdt
dC ),(
![Page 30: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/30.jpg)
Controlling -max
0
( )Lg C ds
I g
Smoothness
Cohen Kimmel, IJCV 97
![Page 31: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/31.jpg)
Fermat’s PrincipleIn an isotropic medium, the paths taken by
light rays are extremal geodesics w.r.t.
i.e.,
Cohen Kimmel, IJCV 97
),(),( yxIyxg
CC
dssCg ))((minarg
![Page 32: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/32.jpg)
Experiments - Color Segmentation
Goldenberg, Kimmel, Rivlin, Rudzsky,
IEEE T-IP 2001
NNCgCgdt
dC
dsCg
),(
)(
![Page 33: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/33.jpg)
Tumor in 3D MRI
Caselles,Kimmel, Sapiro, Sbert, IEEE T-PAMI 97
NNSgHSgdt
dS
daSg
),(
)(
![Page 34: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/34.jpg)
Segmentation in 4D
NNMgHMgdt
dM ),(
Malladi, Kimmel, Adalsteinsson, Caselles, Sapiro, Sethian
SIAM Biomedical workshop 96
NNMgHMgdt
dM
dvMg
),(
)(
![Page 35: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/35.jpg)
Tracking in Color Movies
Goldenberg, Kimmel, Rivlin, Rudzsky,
IEEE T-IP 2001
![Page 36: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/36.jpg)
Tracking in Color Movies
Goldenberg, Kimmel, Rivlin, Rudzsky,
IEEE T-IP 2001
![Page 37: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/37.jpg)
Edge Gradient Estimators
)],(),,([),( yxvyxuIyxI
Xu-Prince 98, Paragios et al. 01, Vasilevskiy-Siddiqi 01, Kimmel-Bruckstein 01
![Page 38: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/38.jpg)
Edge Gradient Estimators
We want a curve with large points and small ‘s so:
Consider the functional
Where is a scalar function, e.g. .
C
dsINCE ),()(
)],(),,([ yxvyxuI
)(sC
)(sN
cos|)(|)(, CICIN
)(s
|| I
) ( ) (
![Page 39: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/39.jpg)
The Classic ConnectionSuppose and we consider a closed contour for C(s).We have
and by Green’s Theorem we have
dsxyvudsNICEsC
ss
sC
)()(
),(),,(,)(
dxdyyxI
dxdyII
Ivu
dxdyuvCE
sC
sC
yyxx
sC
xy
)( nArea withi
)( nArea withi
)( nArea withi
),( of Laplacian
yields this,),(for But,
)(
)(
![Page 40: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/40.jpg)
Therefore:
Hence curves that maximize are curves that enclose all regions where is positive!
We have that the optimal curves in this case are The Zero Crossings of the Laplacian
isn’t this familiar?
The Classic Connection
dxdyIGCEC
)*()(
),(* yxIG
C )(sC
IG *)(CE
IG *
![Page 41: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/41.jpg)
It is pedagogically nice, but the MARR-HILDRETH edge detector is a bit too sensitive.
So we do not propose a grand return to MH but a rethinking of the functionals used in active contours in view of this.
INDEED, why should we ignore the gradient directions (estimates) and have every edge integrator controlled by the local gradient intensity alone?
The Classic Connection
![Page 42: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/42.jpg)
Our Proposal
Consider functional of the form
These functionals yield “regularized” curves that combine the good properties of LZC’s where precise border following is needed, with the good properties of the GAC over noisy regions!
s...other term ||1
1)( where
))((),()(
ICg
dssCgdsINCECC
![Page 43: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/43.jpg)
Implementation Details
We implement curve evolution that do gradient descent w.r.t. the functional
Here the Euler Lagrange Equations provide the explicit formulae.
For closed contours we compute the evolved curve via the Osher-Sethian “miracle” numeric level set formulation.
C
tCE
dt
dC
))((
)(CE
![Page 44: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/44.jpg)
Closed contours
EL eq.
0,)div(),sign( NgNgVVN
GAC
GAC
LZC
LZC
Kimmel-Bruckstein IVCNZ01
gLL
![Page 45: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/45.jpg)
Closed contours
EL eq.
0,)div(),sign( NgNgVVN
GAC
LZC
LZC+GAC
Kimmel-Bruckstein IVCNZ01
gLL
![Page 46: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/46.jpg)
Along the curveb.c. at C(0) and C(L)
Open contours 0)div(),sign( NVVN
0,),sign(|),|1( NVTVNTVN
Kimmel-Bruckstein IVCNZ01
LL
![Page 47: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/47.jpg)
Open contours
Kimmel-Bruckstein IVCNZ01
LL
![Page 48: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/48.jpg)
Geometric Measures
Weighted arc-length
Weighted area
Alignment
Robustalignment e.g.
RR 2:, yx
0),(
Ngdayxg
0),())(( NNggdssCgC
0)div(, NVdsVNC
0)div(),sign(|,| NVVNdsVNC
0, NIdsINC
Variational meaning for Marr-Hildreth edge detector Kimmel-Bruckstein IVCNZ01
![Page 49: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/49.jpg)
Geometric Measures
Minimal varianceChan-Vese, Mumford-Shah, Max-Lloyd, Threshold,…
0))((
)()(
221
\
22
21
21
NIcc
dacIdacI
cc
CC
C
0))(( 212t
2
1
21
\
\
NIccC
c
c
cc
da
Ida
da
Ida
C
C
C
C
C
C \
![Page 50: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/50.jpg)
Geometric Measures
Robust minimal deviation
0- 21
\
21
NcIcI
dacIdacICC
C
NcIcIC
yxIc
yxIc
C
C
12t
\2
1
),(median
),(median
C
C \
![Page 51: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/51.jpg)
Haralick/Canny-like Edge Detector
Haralick suggested as edge detector0I
IIIII yyxx
III
Laplace
Alignment Topological Homogeneity
![Page 52: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/52.jpg)
Haralick/Canny Edge Detector
III
II I
div
III
dIdsdIdxdyIdxdyICCC
II 2
0I
Haralick
co-area
h hdxdyIC
2
x
y
I
Thus, indicates optimal alignment + topological homogeneity
0I
![Page 53: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/53.jpg)
Closed Contours & Level Set Method
implicit representation of CThen,
Geodesic active contour level set formulation
Including weighted (by g) area minimization
RR 2:, yx 0),(:, yxyxC
dC d
VN Vdt dt
yxgdt
d,div
divd
g gdt
y
x
C(t)
C(t) level set 0
, ,x y t
x
y
![Page 54: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/54.jpg)
Operator Splitting Schemes
Additive operator splitting (AOS) Lu et al. 90, Weickert, et al. 98 unconditionally stable for non-linear diffusion
Given the evolution write
Consider the operator Explicit scheme
, the time step, is upper bounded for stability
gt div 00 uu
ll xxl yxgA ,
k
ll
k A
2
1
1 Ι
2
1
divl
xx llgg
![Page 55: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/55.jpg)
LOD:
Operator Splitting Schemes
Implicit scheme
inverting large bandwidth matrix First order, semi-implicit, additive operator splitting (AOS),
or locally one-dimensional (LOD) multiplicative schemes are
stable and efficient given by linear tridiagonal systems of equations
that can be solved for by Thomas algorithm
k
ll
k A 12
1
1
I
k
ll
k A
2
1
11 22
1I
1k
kl
l
k A 12
1
1
I AOS:
![Page 56: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/56.jpg)
Operator Splitting Schemes
We used the following relation (AOS)
Locally One-Dimensional scheme (LOD)
Decoupling the axes and the implicit formulation leads to computational efficiencyThe 1st order `splitting’ idea is based on the operator expansion
)(21212
1 )1( 21
21
11
211 OAAAA kkkk
)(11 )1( 212
11
121
1 OAAAA kkkk
)(11 21 OAA
![Page 57: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/57.jpg)
The geodesic active contour model
Where I is the image and the implicit representation of the curve
If is a distance, then , and the short time evolution is
Note that and thus can be computed once for the whole image
1
Igt div
IAl IAlI
, ,x y t
Example: Geodesic Active Contour
y
x
C(t)
x
y
Goldenberg, Kimmel, Rivlin, Rudzsky,
IEEE T-IP 2001
),(div yxIgt
![Page 58: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/58.jpg)
Example: Geodesic Active Contour
is restricted to be a distance map:Re-initialization by Sethian’s fast marching method every iteration in O(n).
Computations are performed in a narrow band around the zero set
Multi-scale approach:process a Gaussian pyramid of the image
NNO log
tyxz ,,
y
x
C(t)
x
y
![Page 59: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/59.jpg)
Tracking Objects in Movies
Movie volume as a spatial-temporal 3D hybrid space The AOS scheme is
Edge function derived by the Beltrami framework Sochen Kimmel Malladi 98
Contour in frame n is the initial condition for frame n+1.
k
ll
k IA 11 33
1 I
222
222
222
1
1
1
BGRBBGGRRBBGGRR
BBGGRRBGRBBGGRR
BBGGRRBBGGRRBGR
g
yyyxxx
yyyyyyyxyxyx
xxxyxyxyxxxx
ij
x
x
y
y
t
t
![Page 60: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/60.jpg)
Experiments - Curvature Flow
20 50
![Page 61: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/61.jpg)
Experiments - Curvature Flow CPU Time
400
600
800
1000
1200
0 1 2 3 4 5 6 7 8 9 10 11
time step
CP
U t
ime
(sec
) [1
]
0
50
100
150
200
250
300
CP
U t
ime
(sec
) [2
,3]
1. explicit (w hole image) 2. explicit (narrow ) 3. AOS (narrow )
![Page 62: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/62.jpg)
Tracking
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 63: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/63.jpg)
Tracking
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 64: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/64.jpg)
Tracking
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 65: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/65.jpg)
Information extraction
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 66: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/66.jpg)
Holzman-Gazit, Goldshier, Kimmel 2003
=
I I I I
I H I
Thin Structures
![Page 67: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/67.jpg)
Segmentation in 3D
Change in topology
Caselles,Kimmel, Sapiro, Sbert, IEEE T-PAMI 97
![Page 68: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/68.jpg)
Gray Matter Segmentation
Goldenberg Kimmel Rivlin Rudzsky, VLSM 2001
Coupled surfaces
EL equations
dxdydzhg
dxdydzhg
||)(
||)(
212
121
),(||
div
),(||
div
1222
222
2111
111
Fg
Fg
t
t
![Page 69: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/69.jpg)
Gray Matter Segmentation
Goldenberg Kimmel Rivlin Rudzsky, VLSM 2001
![Page 70: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/70.jpg)
Gray Matter Segmentation
Goldenberg Kimmel Rivlin Rudzsky, VLSM 2001
![Page 71: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/71.jpg)
![Page 72: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/72.jpg)
Classification (dogs & cats)
walk run gallop cat...
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 73: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/73.jpg)
![Page 74: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/74.jpg)
Classification (people)
walk run run45
Goldenberg, Kimmel, Rivlin, Rudzsky, ECCV 2002
![Page 75: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/75.jpg)
![Page 76: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/76.jpg)
Conclusions Geometric-Variational method for segmentation
and tracking in finite dimensions based on prior knowledge (more accurately, good initial conditions).
Using the directional information for edge integration.
Geometric-variational meaning for the Marr-Hildreth and the Haralick (Canny) edge detectors, leads to ways to design improved ones.
Efficient numerical implementation for active contours.
Various medical and more general applications.www.cs.technion.ac.il/
~ron
![Page 77: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/77.jpg)
Edge Indicator Function for Color
Beltrami framework: Color image = 2D surface in space The induced metric tensor for the image surface
Edge indicator = largest eigenvalue of the structure tensor metric. It represents the direction of maximal change in
, , , , , , ,GRx y x y y Bx x y , , , ,Rx Gy B
i i j
ji
i
ii uuuu2
222
2
1
2
1
2
11
2 2 2
2 2 2
1
1x x x x y x y x y
ijx y x y x y y y y
R R R
R R
G G GB B B
B B BG GRGg
2 2 2Gd d BR d
X
I
Y
![Page 78: Geometric Active Contours Ron Kimmel ron Computer Science Department Technion-Israel Institute of Technology Geometric Image Processing](https://reader036.vdocuments.net/reader036/viewer/2022062308/56649d1c5503460f949f1859/html5/thumbnails/78.jpg)
AOS
Proof:
The whole low order splitting idea is based on the operator expansion
)(1 )()(11
)()(221
11
)(221
)()(1
)(221
)(1
)(221
)(22
2
1
)(221
2121
2
1
21
1
21
1
2
1
221
221
22
2121
221
221
221
21
221
212
21
21
21
OAAOOAA
OOAA
AA
OAA
OAA
OAA
AA
OAA
AA
OAA
AA
AA
)(21
1
21
1
2
11 2
2121
O
AAAA
)(11
1 2
OAA