Introduction to color science

n
n
n
n
n
n
n

Trichromacy"
Spectral matching functions"
CIE XYZ color system"
xy-chromaticity diagram"
Color gamut"
Color temperature"
Color balancing algorithms"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 1

Newtons Prism Experiment - 1666

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 2

Color: visible range of the electromagnetic spectrum

380 nm

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 3

760 nm

Radiometric Quantities

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 4

Photometric Quantities

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 5

Human retina

[Roorda, Williams, 1999]!

Pseudo-color image of nasal retina,

1 degree eccentricity, in two male subjects, scale bar 5 micron"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 6

Absorption of light in the cones of the human retina

535 nm 575 nm
445 nm

Sensitivity

: ()"
: ()"
: ()
Note: curves are
normalized. Much lower
sensitivity to Blue, since
fewer S-cones absorb
less light."

Wavelength (nm)
Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 7

Three-receptor model of color perception

S ( )C ( ) d

S ( ) C ( ) d

S ( ) C ( ) d

C ( )
Spectral energy distribution"
of incident light"

Effective cone stimulation"

(tristimulus values)"

[T. Young, 1802] [J.C. Maxwell, 1890]!

Different spectra can map into the same tristimulus values

and hence look identical (metamers) "
Three numbers suffice to represent any color"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 8

Color matching
n
n
n

Suppose 3 primary light sources with spectra Pk(), k =1,2,3

Intensity of each light source can be adjusted by factor k"
How to choose k, k =1,2,3, such that desired tristimulus values (R, G, B) result ?"

( )

C =

( )
+ P ( )
+ P ( )
1 P1
2 2

3 3

S ( ) C ( ) d
R

( )

( )

( )

S ( ) C ( ) d

S ( ) C ( ) d

= 1 K i,1 + 2 K i,2 + 3 K i,3

( ) ( )

with K i, j = Si Pj d

Color matching is linear!!

"
Digital
" Image Processing: Bernd Girod, 2013 Stanford University

-- Color 10

( )

i = Si %& 1 P1 + 2 P2 + 3 P3 '( d

Additive vs. subtractive color mixing

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 11

Color matching experiment

Eyes

Courtesy B. Wandell, from [Foundations of Vision, 1996]"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 12

Spectral matching functions

RGB Color Matching Functions
"
"

Tristimulus values

()"
()"
()

n
n

Wavelength (nm)
435.8 nm 546.1 nm

700.0 nm

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 13

Color matching experiment: Monochromatic test

light and monochromatic primary lights"
Spectral RGB primaries (scaled, such that
R=G=B matches spectrally flat white)."
Negative intensity: color is added to test color"
Standard human observer: CIE (Commision
Internationale de LEclairage), 1931."

Luminosity function
Monochromatic"
test light"

reference"
light"

Luminous efficiency

Peak 555nm

Wavelength (nm)

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 15

Experiment:
Match the brightness of a white
reference light and a monochromatic
test light of wavelength "
Links photometric to radiometric
quantities"

CIE 1931 XYZ color system

Tristimulus values

XYZ Color Matching Functions

"
"

Properties:"
n All positive spectral matching functions"

X .490 .310 .200 R

Y
=
.177
.813
.011

G
Z .000 .010 .990 B


n
n
n

Wavelength (nm)

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 16

Y corresponds to luminance"
Equal energy white: X=Y=Z "
Virtual primaries"

Color gamut and chromaticity

Z

X+Y+Z=1
X
Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 18

X
x=
X +Y + Z
Y
y=
X +Y + Z

CIE chromaticity diagram

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 19

Perceptual non-uniformity of xy chromaticity

Just noticeable chromaticity
differences (10X enlarged)"

[MacAdam, 1942]!

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 20

Color gamut
NTSC phosphors
"
R: x=0.67, y=0.33"
G: x=0.21, y=0.71"
B: x=0.14, y=0.08"
"
Reference white:
x=0.31, y=0.32
Illuminant C"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 21

White at different color temperatures

Tc(K)
10000
25000

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 23

3000 2500

4000
6000

2000
1500

Blackbody radiation
Plancks Law Curves"
7000 K"
5770 K"
4000 K

BT() (W m-2 sr-1 Hz-1)

414 nm"

Plancks Law, 1900!

2hc 2 / 5
BT ( ) = hc/ kT
e
1
Wiens Law !

2,900,000
peak [nm] =
T[K]

503 nm"

725 nm"
Wavelength (nm)"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 24

Color balancing
n
n
n
n

Effect of different illuminants can be cancelled only in the spectral domain (impractical)"
Color balancing in 3-d color space is practical approximation "
Color constancy in human visual system: gain control in cone space LMS [von Kries,1902]!
Von Kries hypothesis applied to image acquisition devices (cameras, scanners)"

kL
L

Camera
sensors"

3x3 kM
matrix" M
T"
1

kS

How to determine kL, kM, kS automatically?"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 25

3x3
M matrix"
T"
S

Color balancing (cont.)

n

Von Kries hypothesis"

L k L

M = 0
S 0

0
kM
0

0 L

0 M

kS
S

If illumination (or a patch of white in the scene) is known, calculate"

Ldesired
M desired
S desired
kL =
; kM =
; kS =
Lactual
M actual
Sactual

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 26

Color balancing with unknown illumination

n

k L L x, y = k M M x, y =kS S x, y

Gray-world"

x, y

Scale-by-max"

x, y

Shades-of-gray

[Finlayson, Trezzi, 2004]!

"
"

x, y

k L max L !" x, y #$ = k M max M !" x, y #$ = kS max S !" x, y #$

"
"
n

x, y

x, y

1
p

x, y

1
p

p
p
p
k L L x, y = k M M x, y = kS S x, y
x, y

x, y

x, y

Special cases: gray-world (p = 1), scale-by-max ( p = )"

Best performance for p 6
n

Refinements:
smooth image, exclude saturated color/dark pixels,
use spatial derivatives instead (gray-edge, max-edge)
[van de Weijer, 2007])"
"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 27

1
p

Color balancing example

Original"

Gray-world" Scale-by-max" Gray-edge"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 28

Max-edge" Shades-of-gray"

Color balancing example

Original"

Gray-world"

Scale-by-max"

Gray-edge"

Max-edge"

Shades-of-gray"

Original image courtesy Ciurea and Funt"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 29

Daylight D65"
CIE observer"

Daylight D65"
cheap camera"

Illuminant A"
CIE observer"

Digital Image Processing: Bernd Girod, 2013 Stanford University -- Color 31