Download Week 10 (4/20/2006)

Document related concepts

F-number wikipedia , lookup

Nonlinear optics wikipedia , lookup

Aperture wikipedia , lookup

Harold Hopkins (physicist) wikipedia , lookup

3D optical data storage wikipedia , lookup

Photon scanning microscopy wikipedia , lookup

Optical aberration wikipedia , lookup

Optical coherence tomography wikipedia , lookup

Silicon photonics wikipedia , lookup

Optical tweezers wikipedia , lookup

Transcript
Optical Flow
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Optical Flow:
Where do pixels move to?
Motion is a basic cue
Motion can be the only cue for segmentation
Applications







tracking
structure from motion
motion segmentation
stabilization
compression
mosaicing
…
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Definition of optical flow
OPTICAL FLOW = apparent motion of
brightness patterns
Ideally, the optical flow is the projection of the
three-dimensional velocity vectors on the image

Caution required !
Two examples :
1. Uniform, rotating sphere

O.F. = 0
2. No motion, but changing lighting

O.F.  0

Mathematical formulation
I (x,y,t) = brightness at (x,y) at time t
Brightness constancy assumption:
dx
dy
I ( x  t , y  t , t  t )  I ( x, y, t )
dt
dt
Optical flow constraint equation :
dI I dx I dy I


 0
dt x dt y dt t

Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
The aperture problem
dx
u  
dt
I
I x  
y
dy
v
dt
I
I y  
y
I
It 
t
I xu  I y v  I t  0
1 equation in 2 unknowns

The Aperture Problem
0
What is Optical Flow, anyway?



Estimate of observed projected motion field
Not always well defined!
Compare:

Motion Field (or Scene Flow)
projection of 3-D motion field

Normal Flow
observed tangent motion

Optical Flow
apparent motion of the brightness pattern
(hopefully equal to motion field)

Consider Barber pole illusion
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Horn & Schunck algorithm
Additional smoothness constraint :
es   (( ux2  u 2y )  (vx2  v 2y )) dxdy,
besides OF constraint equation term
ec   ( I x u  I y v  I t )2 dxdy,
minimize es+ec

Horn & Schunck
The Euler-Lagrange equations :


Fu  Fux  Fu y  0
x
y


Fv  Fvx  Fv y  0
x
y
In our case ,
F  (u  u )  (v  v )   ( I xu  I y v  I t ) ,
2
x
2
y
2
x
2
y
so the Euler-Lagrange equations are
u   ( I x u  I y v  I t ) I x ,
v   ( I xu  I y v  I t ) I y ,

2
2
 2  2
x
y
is the Laplacian operator
2
Horn & Schunck
Remarks :
1. Coupled PDEs solved using iterative
methods and finite differences
u
 u   ( I x u  I y v  I t ) I x ,
t
v
 v   ( I x u  I y v  I t ) I y ,
t
2. More than two frames allow a better
estimation of It
3. Information spreads from corner-type
patterns


Horn & Schunck, remarks
1. Errors at boundaries
2. Example of regularization
(selection principle for the solution of
ill-posed problems)

Results of an enhanced system
Structure from motion with OF
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
dE (u, v)
  2 I x I xu  I y v  I t   0
du
dE (u, v)
  2 I y I xu  I y v  I t   0
dv
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Bayesian flow
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Optical Flow










Brightness Constancy
The Aperture problem
Regularization
Lucas-Kanade
Coarse-to-fine
Parametric motion models
Direct depth
SSD tracking
Robust flow
Textured motion
Optical Snow


http://www.cim.mcgill.ca/~langer/research-opticalsnow.html
Optical snow is the type of motion an observer sees
when watching a snow fall. Flakes that are closer to
the observer appear to move faster than flakes which
are farther away.
Textured Motion


http://civs.stat.ucla.edu/Yizhou_Research/Tex
turedmotion.htm
Natural scenes contain rich stochastic motion
patterns which are characterized by the
movement of a large amount of particle and
wave elements, such as falling snow, water
waves, dancing grass, etc. We call these
motion patterns "textured motion".