W10: Colour

Course Overview

• Course Title: CSCI 3090
• Focus: Colour
• Instructor: Shima Rezasoltani, Faculty of Science, Ontario Tech

Goals

• By the end of the class, students will:
  • Have a basic understanding of colour vision
  • Understand basic colour spaces
  • Grasp issues in using colour for data representation

Introduction to Colour

• Understanding colour is crucial in advanced computer graphics and visualization.
• Colour involves:
  • Physics
  • Physiology of human vision
  • Psychology of colour perception

Human Vision Systems

• Two types of vision systems in humans:
  1. Colour Vision System:
  • Functions under normal lighting conditions
  • Main focus in this section
  1. Black and White Vision System:
  • Operates under low light
  • Measures light intensity

How Vision Systems Work

• Vision systems detect light and provide a single value for it:
  • Human vision can detect wavelengths from 380 nm to 800 nm.
  • Detectors compute a weighted average over these wavelengths.

Detector Response Equation

• The response is characterized by: $R = ext{integral} \big( L( heta) w( heta) d heta$ where
  • $heta$ is wavelength
  • $L( heta)$ is light amount at $heta$
  • $w( heta)$ is the weighting function
• Weighting functions are experimentally determined and tabulated.

Cones in Human Colour Vision

• Human colour vision relies on three types of cone detectors:
  • Each cone has an associated $w( heta)$ characteristic, showing overlap in response.
  • Different combinations of signals can produce the same perceived colour.

Metamers

• Two different spectra can produce the same colour (metamers), useful in production for visual consistency despite ease of production.

Challenges in Colour Perception

• Since colour perception is both perceptual and physical, developing a consistent mathematical theory is challenging.
• Early methods in the 1980s used indirect techniques to study colour vision using overlapping coloured lights.

Colour Matching Techniques

• Colour matching used independent dimmers for red, green, and blue lights to mix and match various colours.
• Some colours cannot be exactly matched due to limitations in physical colours.

CIE Colour Standard

• In the 1930s, the CIE established a standard:
  • Primaries were designated as X, Y, and Z.
  • $Y$ is related to brightness, while $X$ and $Z$ relate to hue.

Projects in CIE Colour Space

• CIE colour space uses two coordinates (x, y) after projecting onto the X+Y+Z=1 plane where $z = 1 - x - y$.
• This results in removing luminance, allowing hue to be assessed.

Gamut Representation

• A plot representing the range of real colours corresponds to the potential colours in a specific colour space (gamut).
• The colour gamut shows limitations of colour representation within RGB and XYZ spaces.

Non-Linear Colour Spaces

• RGB and XYZ colour spaces are not perceptually linear—differences may appear distorted across colour spaces.
• This non-linearity can lead to misleading interpretations of visualized data.

Desirable Colour Spaces

• A perceptually linear colour space is desirable, where differences correspond linearly to perceived differences.
• The Luv* and Lab* are close approximations.

Conversion Between Colour Spaces

• Reference white colour ($X<em>n, Y</em>n, Z_n$) is scaled to 100 in Luv* and Lab*.
• Transformation can be done between Luv, Lab and XYZ spaces using specific formulas.

Practical Applications

• Matrices are used to transform between RGB and XYZ colour spaces.
• Different devices (monitors and printers) have distinct gamuts, creating challenges in colour consistency.

Summary

• Colour theory is complex; caution is needed in visualization to avoid misleading interpretations.
• Device calibration is crucial for consistent colour representation across different platforms.