Lecture Three Colour Models

1
Lecture ThreeColour Models
2
Overview

• Foundations of light and colour
• Colour representation
• RGB colour cube
• Other hardware-oriented colour models
• Limitations of hardware-oriented representations
• HSV and user-oriented colour models

3
Light

• A narrow frequency band within the spectrum of
  electromagnetic energy
• the visible spectrum from 400-700nm
• others are ultraviolet, infrared, microwaves etc.
• each wavelength within the visible spectrum
  produces light of a different colour
• 400nmviolets, blues, greens, yellows,
  700nmoranges, reds

4
Spectral Density

• P(?) - The power per unit wavelength of a
  coloured light

P(?)
? (nm)
400
700
Red
Blue
violet
Indigo
Green
Yellow
Orange
5
Colour Description

• Coloured light is described in terms of
• Hue the perceived colour (red/yellow etc)
  determined by the dominant wavelength. No
  dominant wavelength achromatic
• Saturation the purity of the colour
• Brightness/Luminance the perceived intensity of
  the light
• The Chrominance of a colour is the combination of
  hue and saturation

6
Dominant Wavelength Model
P(?)
B
D
A
??(nm)
700
400
550
W
Dominant wavelength _at_ 550nm green Luminance (L)
(D-A)B AW Saturation (D-A)B (100/L)
7
Colour Definition

• Dominant wavelength method
• useful for description of colours, but not for
  precisely obtaining and representing colour
  values
• A more precise method is based on the reception
  of coloured light by the human eye

8
The Human Visual System

• Retina
• light-sensitive membrane consisting of three
  types of colour sensors (cones)
• each type is most receptive to wavelengths in
  either the red, green or blue ranges
• perceived colour is a result of the relative
  excitation of each group of cones
• Leads to a 3-D representation of colour based on
  red, green blue primaries

9
RGB Colour Cube
Green
Green (0,1,0)
Yellow (1,1,0)
White (1,1,1)
Cyan (0,1,1)
Black (0,0,0)
Red
Red (1,0,0)
Blue (0,0,1)
Magenta (1,0,1)
Blue
10
Using RGB

• Hardware-oriented model
• Equally defined independent RGB values are well
  suited to graphics architecture
• Greyscales along line where RGB
• Additive colour specification
• colours defined in terms of an addition to black
• Linear colour combination operation
• suitable for colour merging operations, but less
  useful for colour mixing

11
Other Hardware Oriented Models

• CMYK
• Cyan, Magenta, Yellow, Black
• Subtractive primaries
• colour is specified as a subtraction from white
• used in printing industry
• YUV, YIQ
• Broadcast standards (YUVUK, YIQUSA)
• YLuminance, UV/IQ are chrominance
• RGB re-coded for narrow transmission bandwidth

12
User-Oriented Models

• RGB-based models are derived from a good fit with
  hardware requirements
• but they do not provide an intuitive means of
  user colour specification
• e.g. how to specify brown, gold, etc.?
• User-oriented models attempt to view colour using
  the perceptive terms identified earlier

13
HSV (Hue, Saturation, Value)

• Developed by AR Smith (1978)
• re-coding of RGB colour cube to generate a
  user-view of colour
• set RGB cube on black vertex and look down from
  the white vertex
• primary and secondary colours are arranged
  radially around the centre axis

14
The HSV Hexicone
Green
Yellow
White
Cyan
Red
Blue
Magenta
Value
Hue
Black
Saturation
15
Using HSV

• More intuitive colour selection
• First, select the pure hue which most closely
  matches the colour
• Lighten the colour by adding white (saturation)
• Darken the colour by adding black (value)
• Note
• conversion needed to RGB for display (exercise)
• value is not an exact model of brightness