Any colour that can be specified using a model will correspond to a single point within the subspace it defines. Each colour model is oriented towards either specific hardware (RGB,CMY,YIQ), or image processing applications (HSI).

1. COLOR IMAGE
PROCESSING

2. Roadmap
2
Image
Acquisition
Image
Enhancement
Image
Restoration
Image
Compression
Image
Segmentation
Representation
& Description
Recognition &
Interpretation
Knowledge Base
Preprocessing – low level
Image
Coding
Morphological
Image Processing
Wavelet
Analysis

3. Color spectrum
3
 When passing through a prism, a beam of sunlight is
decomposed into a spectrum of colors: violet, blue,
green, yellow, orange, red
 1666, Sir Isaac Newton

4. Electromagnetic energy spectrum
4
 Ultraviolet  visible light  infrared
 The longer the wavelength (meter), the lower the frequency (Hz), and
the lower the energy (electron volts)
 The discovery of infrared (1800, Sir Frederick William Herschel)
 What is infrared?
http://coolcosmos.ipac.caltech.edu/cosmic_classroom/ir_tutorial/

5. Hyperspectral imaging
5
 AVIRIS (Airborne Visible-Infrared Imaging Spectrometer)
 Number of bands: 224
 Wavelength range (mm): 0.4-2.5
 Image size: 512 x 614
 Spectral range
 visible light (0.4 ~ 0.77mm)
 near infrared (0.77 ~ 1.5mm)
 medium infrared (1.5 ~ 6mm)
 far infrared (6 ~ 40mm)

6. Some questions
6
 What does it mean when we say an object is in a
certain color?
 Why are the primary colors of human vision red,
green, and blue?
 Is it true that different portions of red, green, and
blue can produce all the visible color?
 What kind of color model is the most suitable one to
describe human vision?

7. Primary colors of human vision
7
 Cones are divided into three sensible
categories
 65% of cones are sensitive to red light
 33% are sensitive to green light
 2% are sensitive to blue light
 For this reason, red, green, and blue are
referred to as the primary colors of
human vision. CIE standard designated
three specific wavelength to these three
colors in 1931.
 Red (R) = 700 nm
 Green (G) = 546.1 nm
 Blue (B) = 435.8 nm
Detailed experimental
Curve available in 1965
Detailed experimental
curve available in 1965

8. Some clarifications
8
 No single color may be called red, green, or blue.
 R, G, B are only specified by standard.

9. Secondary colors
9
 Magenta (R + B)
 Cyan (G + B)
 Yellow (R + G)

10. Primary colors of pigment
10
 A primary color of pigment refers to one
that absorbs the primary color of the light,
but reflects the other two.
 Primary color of pigments are magenta,
cyan, and yellow
 Secondary color of pigments are then red,
green, and blue

11. 11

12. Additive vs. Subtractive color system
 involves light emitted directly
from a source
 mixes various amounts of red,
green and blue light to produce
other colors.
 Combining one of these
additive primary colors with
another produces the additive
secondary colors cyan,
magenta, yellow.
 Combining all three primary
colors produces white.
 Subtractive color starts with an
object that reflects light and
uses colorants to subtract
portions of the white light
illuminating an object to
produce other colors.
 If an object reflects all the
white light back to the viewer, it
appears white.
 If an object absorbs (subtracts)
all the light illuminating it, it
appears black.
12

13. Color characterization
13
 Brightness: chromatic notion of intensity
 Hue: dominant color perceived by an observer
 Saturation: relative purity or the amount of
white mixed with a hue
R
G
B
H
S
0o
120o
240o

14. Some clarifications
14
 So when we call an object red, orange, etc. we
refer to its hue

15. Chromaticity
15
 Chromaticity: hue +
saturation
 Tristimulus: the amount
of R, G, B needed to
form any color (X, Y, Z)
 Trichromatic coefficients:
x, y, z
1=++
++
=
++
=
++
=
zyx
ZYX
Z
z
ZYX
Y
y
ZYX
X
x

16. Chromaticity diagram
16
Specifying colours systematically can be achieved using
the CIE chromacity diagram
On this diagram the x-axis represents the proportion of
red and the y-axis represents the proportion of red
used
The proportion of blue used in a colour is calculated as:
z = 1 – (x + y)

17. Chromaticity diagram
17

18. This means the entire
colour range cannot be
displayed based on any
three colours
The triangle shows the
typical colour gamut
produced by RGB
monitors
The strange shape is the
gamut achieved by high
quality colour printers
18

19. Color models
19
 RGB model
 Color monitor, color video cameras
 CMY model
 Color printers
 HSI model
 Color image manipulation

20. RGB model
20
 Color monitor, color video cameras
(additive color system)
 Pixel depth – nr of bits used to represent
each pixel
 Full color image (24 bits)

21. RGB
21

22. CMY model
22
 Color printers and copiers (subtractive color system)
 CMYK color model
 Four color printing
 Deposit colored pigment on paper
ú
ú
ú
û
ù
ê
ê
ê
ë
é
-
ú
ú
ú
û
ù
ê
ê
ê
ë
é
=
ú
ú
ú
û
ù
ê
ê
ê
ë
é
B
G
R
Y
M
C
1
1
1

23. HSI model
23
 The intensity component (I) is decoupled from the
color components (H and S)
 Ideal for developing image processing algorithms
 H and S are closely related to the way human visual
system perceives colors

24. Hue and Saturation
24

25. Hue, Saturation, Intensity
25

26. RGB-to-HSI conversion
26
 
 
    
    






















BG
BG
H
BGBRGR
BRGR
BGR
I
S
BGRI
2
2
1
cos
],,[min
3
1
3
1
2
1



Given a color as R, G, and B its H, S, and I values are
calculated as follows

27. HSI-to-RGB conversion (*)
27
 Given a color as H, S, and I it’s R, G, and B values are
calculated as follows:
 -RG sector (0o <= H < 120o)
 -GB sector 120o <= H < 240o
 For 240o <= H < 360o
 
    )(3,1,
60cos
cos
1 BRIGSIB
H
HS
IR 






 
 
    )(3,1,
180cos
120cos
1 GRIBSIR
H
HS
IG 







 

 
    )(3,1,
300cos
240cos
1 BGIRSIG
H
HS
IB 







 
