Beginner’s guide to CIE colour spaces in Computer Vision

Colour spaces serve as mathematical models that offer a means to express colours as numeric values. Various colour spaces are employed to fulfil different objectives, each presenting its distinct approach to depicting colours.

This article includes brief description of five color spaces including

1. RGB
2. XYZ
3. Lab
4. YUV
5. HSB / HSL

RGB Color Space

RGB stands for Red, Green and Blue. It is a linear colour space and any colour can be created represented as linear combination of red, blue and green by just adding their different intensities. This colour space is compatible with electronic displays and digital imaging devices.

CIE XYZ Color Space

It is a linear colour space, meaning it represents colours as a combination of three primary components: X, Y, and Z. These components are based on the spectral sensitivity of the human visual system and describe how the human eye responds to different wavelengths of light. The Y component represents luminance, while X and Z represent the chromaticity information. CIE XYZ quantifies the actual colours, not the perceived colours. The perceived colour difference of two colours can be quite different from the difference of their XYZ values.

CIE Lab colour space

CIE Lab colour model intends to quantify differences perceived in colour. Lab is defined by three components:

• L represents lightness (0–100).
• a represents the colour on the green to red axis (-128 to 128).
• b represents the colour on the blue to yellow axis (-128 to 128).

CIE YUV Color Space

It separates the luminance (Y) and chrominance (U and V) components of an image. Y represents brightness, while U and V represent color information. Some operations can be performed more efficiently performed in YUV colour space where brightness and colour are stored in different image planes.

HSB or HSL colour space

These colour spaces used to represent colours in a way that corresponds more closely to how humans perceive them. In these colour spaces:

• Hue (H) represents the type of colour, typically measured in degrees around a colour wheel (0°-360°).
• Saturation (S) represents the intensity or purity of the colour (0–100%).
• Brightness (B) in HSB or Lightness (L) in HSL represents the perceived lightness of the colour (0–100%).

For detailed overview of colour spaces and their graph you can check this article by Chandler Abraham, this article is very informative and it provides details about colorimetry.

References

E. R Davies, Computer Vision: Principles, Algorithms, Applications, Learning, Academic Press; 5th edition; 2017

D Forsyth, Computer Vision. A Modern Approach