44
Stefano Soatto (c) UCLA Vision Lab 1 Homogeneous representation Points Vectors Transformation representation

Stefano Soatto (c) UCLA Vision Lab 1 Homogeneous representation Points Vectors Transformation representation

  • View
    216

  • Download
    0

Embed Size (px)

Citation preview

Stefano Soatto (c) UCLA Vision Lab 1

Homogeneous representation

Points

Vectors

Transformation

representation

Stefano Soatto (c) UCLA Vision Lab 2

Lecture 4: Image formation

Stefano Soatto (c) UCLA Vision Lab 3

Image Formation

Vision infers world properties form images. How do images depend on these properties? Two key elements

Geometry Radiometry We consider only simple models of these

Stefano Soatto (c) UCLA Vision Lab 4

Image formation (Chapter 3)

Stefano Soatto (c) UCLA Vision Lab 5

Representation of images

Stefano Soatto (c) UCLA Vision Lab 6

Similar triangles <P’F’S’>,<ROF’> and <PSF><QOF>

(

fzz

11

'

1

2' ))(( ffzfz

Stefano Soatto (c) UCLA Vision Lab 7

Pinhole model

Stefano Soatto (c) UCLA Vision Lab 8

Forward pinhole

Stefano Soatto (c) UCLA Vision Lab 9

Distant objects are smaller

(Forsyth & Ponce)

Stefano Soatto (c) UCLA Vision Lab 10

Parallel lines meet

Common to draw image plane in front of the focal point. Moving the image plane merely scales the image.

(Forsyth & Ponce)

Stefano Soatto (c) UCLA Vision Lab 11

Vanishing points

• Each set of parallel lines meets at a different point– The vanishing point for this direction

• Sets of parallel lines on the same plane lead to collinear vanishing points. – The line is called the horizon for that plane

Stefano Soatto (c) UCLA Vision Lab 12

Properties of Projection

Points project to points Lines project to lines Planes project to the whole image or a half image Angles are not preserved Degenerate cases

Line through focal point projects to a point. Plane through focal point projects to line Plane perpendicular to image plane projects to part of

the image (with horizon).

Stefano Soatto (c) UCLA Vision Lab 13

Orthographic projection

yy

xx

'

'

Stefano Soatto (c) UCLA Vision Lab 14

Stefano Soatto (c) UCLA Vision Lab 15

Cameras with Lenses

(Forsyth & Ponce)

Stefano Soatto (c) UCLA Vision Lab 16

Stefano Soatto (c) UCLA Vision Lab 17

Assumptions for thin lens equation Lens surfaces are spherical Incoming light rays make a small angle with

the optical axis The lens thickness is small compared to the

radii of curvature The refractive index is the same for the

media on both sides of the lens

Stefano Soatto (c) UCLA Vision Lab 18

Stefano Soatto (c) UCLA Vision Lab 19

Blur circle

A point at distance z is imaged at point 'z from the lens

fzz

11

'

1

and so

)()()(

'' zzfz

f

fz

fzz

|''|'

zzz

db

Points a t distance are brought into focus at distance z 'z

Thus points at distance will give rise to a blur circle of diameterz

with d the diameter of the lens

-zz’

-zz’

P

P’

Q

Q’db

Stefano Soatto (c) UCLA Vision Lab 20

Interaction of light with matter Absorption Scattering Refraction Reflection Other effects:

Diffraction: deviation of straight propagation in the presence of obstacles

Fluorescence:absorbtion of light of a given wavelength by a fluorescent molecule causes reemission at another wavelength

Stefano Soatto (c) UCLA Vision Lab 21

Refraction

n1, n2: indexes of refraction

Stefano Soatto (c) UCLA Vision Lab 22

Solid Angle

2

0

2d

radian

2

]cos[2

sin2

sin

0

2

0

2

0

2

0

2

0

d

ddd

steradian (sr)

ddsind

hemisphere

Sphere: 4

Stefano Soatto (c) UCLA Vision Lab 23

Radiometric Terms

Stefano Soatto (c) UCLA Vision Lab 24

Irradiance and Radiance

Irradiance Definition: power per unit area incident on a surface

W/m2 = luxdA

dE

ddA

dL

cos

2

iL

i

A

Radiance Definition: power per unit area and projected solid angle

W/m2sr

iii dLdA

ddE cos

2

2H

iii dLE cos

Stefano Soatto (c) UCLA Vision Lab 25

Radiant Intensity

Radiant flux W

Definition: flux per unit solid angle

W/sr = cd (candela)

d

dI

2S

dI

I

[ ]

Stefano Soatto (c) UCLA Vision Lab 26

Isotropic Point Source

Stefano Soatto (c) UCLA Vision Lab 27

Isotropic Point Source

Radiant flux WAll directions: solid angle 4

Radiant flux per

unit solid angle W/sr

d

dI

Radiant intensity

4

I

r

drdA 2

22 4

1

44 rdA

dA

rdA

d

dA

Id

dA

dE

• Note inverse square law fall off.

Stefano Soatto (c) UCLA Vision Lab 28

Isotropic Point Source

Radiant flux WAll directions: solid angle 4

Radiant flux per

unit solid angle W/sr

d

dI

Radiant intensity

4

I

r

drdAcos

12

2

3

2 444 hdA

dA

rdA

d

dA

Id

dA

dE

coscos

• Note cosine dependency.

h

cosrh

Stefano Soatto (c) UCLA Vision Lab 29

Isotropic Point Source

Point source at a finite distance

r

2

3

4 hE

cos

• Note cosine dependency.

h

• Note inverse square law fall off.

Stefano Soatto (c) UCLA Vision Lab 30

Irradiance from Area Sources

Stefano Soatto (c) UCLA Vision Lab 31

Hemispherical Source

L

dL

dL

ddLE

i

ii

iiii

2

0

2

020

2

2

0

2

0

2

sin2

sincosL

Stefano Soatto (c) UCLA Vision Lab 32

Reflectance

Reflectance: ratio of radiance to irradiance

dLr=fr dEi

Ei

i

Ei

Lr(x,)

Li(x,i)

irr dEfLi

rr dE

dLf

The surface becomes a light source

Stefano Soatto (c) UCLA Vision Lab 33

BRDF

Stefano Soatto (c) UCLA Vision Lab 34

BRDF

i

rr

iii

irriirrriir

d

dd

dE

EdLf

/

,

;,;,,;,

2

iiriririr dEfEdL ;;;

i

iirirrr dEfL

;

ii

irirrir dE

EdLf

;;

;

iiiiiiiii dLdLdE cos

iiiiriri

dLf cos;

Stefano Soatto (c) UCLA Vision Lab 35

Reflection Equation

Stefano Soatto (c) UCLA Vision Lab 36

Perfectly Diffuse Reflectioni

Perfectly Diffuse Surface •Appears equally bright from all viewing directions (r, r)•Reflects all incident light, i.e.,

iiiiBi

dLL cos1

BrrrrBrrr LddLdLEr rr

sincos

LB(r, r) is constant for all directions (r, r)

1

s

BB E

Lf

Stefano Soatto (c) UCLA Vision Lab 37

Common Diffuse Reflectioni

Normal Diffuse Surface •Appears almost equally bright from most viewing directions (r, r), r << 90°•Reflects only a fraction of incident light, i.e.,

iiiiBBi

dLL cos

Bs

BB E

Lf

Reflectance : Albedo

Stefano Soatto (c) UCLA Vision Lab 38

Perfectly Diffuse Reflection

i

Lambertian cosine Law

s

sisiii sin

)()(E),(E

s

iiisisi

iiiiiii

iiiiiBr

E

ddE

ddL

dLfL

i i

i i

i

cos1

cos)()(1

sincos),(1

cos),(

Distant point light source

Stefano Soatto (c) UCLA Vision Lab 39

Law of Reflection

),(),( iiirrr LL

),(L rrr ),(L iii

Stefano Soatto (c) UCLA Vision Lab 40

Perfectly Specular Reflection

),(),( iiirrr LL

ii

ririiirrsf

cossin

)()(),;,(

i i

iiiiiiiiirrsrrr ddLfL

sincos),(),;,(),(

From the definition of BRDF, the surface radiance is:

To satisfy:

i i

iiiiiririrri ddLL

),()()(),(

Stefano Soatto (c) UCLA Vision Lab 41

Lambertian Examples

Scene

(Oren and Nayar)

Lambertian sphere as the light moves.

(Steve Seitz)

Stefano Soatto (c) UCLA Vision Lab 42

Lambertian + Specular Model

)(cos),,()(

cos),,()(),,(

0

00

sn

sss

iiid

PLP

dPLPPL

Stefano Soatto (c) UCLA Vision Lab 43

Lambertian + specular

• Two parameters: how shiny, what kind of shiny.• Advantages

– easy to manipulate– very often quite close true

• Disadvantages– some surfaces are not

• e.g. underside of CD’s, feathers of many birds, blue spots on many marine crustaceans and fish, most rough surfaces, oil films (skin!), wet surfaces

– Generally, very little advantage in modelling behaviour of light at a surface in more detail -- it is quite difficult to understand behaviour of L+S surfaces (but in graphics???)

Stefano Soatto (c) UCLA Vision Lab 44

Lambertian+Specular+Ambient

(http://graphics.cs.ucdavis.edu/GraphicsNotes/Shading/Shading.html)