The Internet belongs to everyone. Let’s keep it that way.

Protect Net Neutrality
Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.



No description

Alla Cordery

on 3 February 2014

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Colour

Digital Images
Subtractive colour
Other Considerations

U08884 Image Technology
Lecture 2

Today's lecture is about:
• Images as an indexable array of pixels
• Sampling and quantisation
• Colour gamut and colour spaces
• Additive and subtractive colour
• Gamma
Chapman, Ch 6
Different conventions
• So, be careful.

• When using a package or programming language to process a picture, you may need to find out what the indexing and origin conventions are for that particular system.

Sampling and quantisation
• When you look round the world, you see what seems like a continuous image. The view does not seem like it is composed of individual pixels.

• When we create a digital image, we have to sample the continuous energy flux onto a discrete grid.

• And, we have to quantize the colour values into a
discrete set of colours.
• Digital colours are also discrete. Not every shade of colour is available.
• To quantize a sampled image, replace each pixel with the nearest in colour to the one that is available.
• But, how do you find the “average colour” when sampling?
• And, how do you know what the “nearest” colour is when quantizing?
• The answer is that colours are represented as numbers.
A useful tutorial on bit depth can be found at:

Grey Scale Range
• The grey scale expresses the brightness or intensity of a pixel between a minimum value and a maximum value.
• There is no colour information.
• Below is a simple grey scale quantised into 8 grey levels.
• Eight levels does not give enough shades of grey.
• A very common range in digital images is 256, so values are numbers ranging from 0 to 255.

A very useful tutorial on digital images can be seen here:
How about colour?
Well, instead of shades of grey, we could assign a colour to a pixel.
How do we do this? Do we simply say 'pink, 'light blue', 'scarlet', 'turquoise'....
No, in reality we assign a number to the pixel which represents the colour that we want.
In order to understand how to assign colours to an image, we need to understand the various colour models and the theory behind it.

The basics - Light and the Spectrum
The basics - Human vision
Red, Green, Blue
Additive colour - R,G,B
Any colour can be described as a combination of Red, Green and Blue light components
RGB space is additive (i.e. additive primaries)
used for TV and computer displays

Does the RGB model represent all colours that we can see?

Actually our eyes are much more capable and they can see colours way beyond what the RGB model provides.

To get a better understanding of this please see the imbeded video.
Other colour space models
Depth resolution:

the number of bits required to store the intensity value
text can be represented with 1-bit resolution
8 bit colour provides 256 different values per colour (sometimes called 24 bit colour - millions of colour)
high quality colour images need 24 bits per colour

Colour depth
HSI (B) - Hue, Saturation, Intensity (Brightness)
Closer to human perception of light
Hue - describes the colour ; i.e. brown (27%c;77%m, 100%y, 18%b) or pink
Saturation - how intense the colour is
Intensity - how much light the colour contains
changing from RGB to HIS and back is easy with the help of a computer

The colour wheel
A red apple is a good example of subtractive color; the apple really has no color; it has no light energy of its own, it merely reflects the wavelengths of white light that cause us to see red and absorbs most of the other wavelengths which evokes the sensation of red. The viewer (or detector) can be the human eye, film in a camera or a light-sensing instrument.
CYM(K) - Cyan, Magenta, Yellow (Black)
Inverse of RGB
Subtractive colour
Used in printing industry
CYM can be easily converted to RGB:
C= G + B = W – R
M = R + B = W – G
Y = R + G = W – B
Complimentary colours

Printing and CYMK gamut

Use of black (K)
Real inks are not perfect, so they do not absorb all of teh complimentary colours
The gamut for CYM(K) is less than RGB
Other colour models
There is no one perfect way to represent colour. Many colour spaces have been designed for different purposes. The main ones are:
• RGB -- Red, Green, Blue
• CMY -- Cyan, Magenta, Yellow
• CMYK -- Cyan Magenta, Yellow, true Black
• HSI -- Hue, Saturation, Intensity
• LAB and LUV -- XYZ models by the Commission Internationale de lʼEclairage (CIE)

HSI (Hue, Saturation, Intensity)
Closer to human perception of light
Hue - describes the colour ; i.e. brown (27%c;77%m, 100%y, 18%b) or pink
Saturation - how intense the colour is
Intensity - how much light the colour contains
changing from RGB to HIS and back is easy with the help of a computer

Considerations for Colour Spaces
• Is it additive or subtractive?

• Is it device independent?

• Is it perceptually uniform?

• Does it separate luminance and chrominanace
No two devices are identical.

A colour that is designed on
your laptop may look
completely different when
projected by a data projector.

Colour management is used to
get round the problem of
device dependence.

Some colour spaces have been
designed to be device
independent: XYZ, sRGB.
sRGB requires each
component to have a gamma
and white point.
A colour space defines its primary components.

• It may also define whether the white point and gamma are to be taken into account. White point and gamma are characteristics of a display device.
• The white point gives an absolute measure of what
“white” means. E.g. a TV monitorʼs “white” is actually quite blue. White point is expressed as a colour temperature in units of Kelvin. Most computer
monitors have 9300K. Daylight (quite yellow) around
7500K, TV about 6500K.
• The gamma describes the relationship between the
value of the colour component and the intensity of the light emitted by the display device.
White point and Gamma

Human vision is non-linear
Digital camera sensor is linear
Computer monitor non-linear
Color temperature describes the spectrum of light which is radiated from a "blackbody" with that surface temperature. A blackbody is an object which absorbs all incident light — neither reflecting it nor allowing it to pass through.
Despite its name, light which may appear white does not necessarily contain an even distribution of colors across the visible spectrum:
Note how 5000 K produces roughly neutral light, whereas 3000 K and 9000 K produce light spectrums which shift to contain more orange and blue wavelengths, respectively.
Suppose we have an image with 24-bit true colour. This means that each R, G, B primary is a number in the range 0-255.

So, a cornflower blue colour might be represented
like this: 133 151 213

• This gives 16,777,216 possible colours. (Why,
because 24 bits is enough to represent any number
between 0 and 16,777,215).

• Each pixel takes 3x8 = 24 bits of memory.

• So, a typical digital screen image of 1280x1024 pixels requires 3145780 bytes (about 3Mb).

Colour Mapping
Colour Mapping
• But suppose we have a very simple image with only a few colours in it, such as a cartoon? We donʼt need 24-bit colour.
• This image has not more than 64 colours.
• So, we could represent this image using a 6-bit pixel to represent one of 64 possible colours (indexed from 0-63).

• If we used a maximum of 256 colours, we would only need 1Mb of memory for a 1280x1024 frame.
• So, we can make a colour map (or color table, or colour palette) to relate numerical colour names (or colour indices) in the range 0-255 to our choice of colours.
• Then when the computer renders the colour on the screen, it looks up the colour in the table, and uses the R, G, B values found there.
Full transcript