HackMD
IMCO: Describing color
Describing a relative color
- Need of a vocabulary to describe a color
- 2 (or more) objects may be all blue, but different
- Need additional descriptions within colors
- How do we organize colors ?
Blue jacket ?
All of this is blue
This is navy blue
This is dark navy blue
What about colors that are hard to describe ?
How ? - Desert Island Experiment
A person with normal color vision is on an island and need to group pebbles
- Think about colo in terms of common color names (red, blue, green, etc.)
- How to arrange achromatic samples ?
- Orders from darker to lighter
- How to arrange chromatics smaples ?
- By color (hue)
-
Example: group all red pebbles together
- How to arrange chromatic samples of the same hue ?
- order by lightness
- order by how much color they contain (chroma)
We know how to arrange chromatic samples of the same hue
Color attributes
We found color attributes
- Pyschological attributes that describe colors:
- Hue
- Lightness
Definition: Hue
Attribute of a visual sensation according to which an area appears to be similar to one of the perceived colors: red, yellow, green, and blue, or to a combination of 2 of them
- achromatic color: perceived with lacking hue
- chromatic color: perceived color with hue
11 primitives colors
\(\color{white}{\text{white}}, \color{gray}{\text{gray, }}\color{black}{\text{black, }} \color{red}{\text{red, }}\color{green}{\text{green, }} \color{yellow}{\text{yellow, }} \color{blue}{\text{blue, }} \color{orange}{\text{orange, }} \color{purple}{\text{purple, }} \color{pink}{\text{pink, }} \color{brown}{\text{brown}}\)
Elementary colors: white, black, red, green, yellow, blue
- unique hues according to opponent color theory
Definition: Lightness
- Brightness: attribute of a visual sensation according to which an area appears to emit, or reflect, more or less light \(\to\) relative
- Lightness: the brightness of an area judged relative to the brightness of a similarity illuminated area that appears to be white or highly transmitting \(\to\) absolute
\[ \text{Lightness}=\frac{\text{Brightness}}{\text{Brightness(white)}} \]
Definition: Chroma
- Colorfulness: Attribute of a visual sensation according to which the perceived color of an area appears to be more or less chromatic
- Chroma: Colorfulness of an area judged as a proportion of the brightness of a similarly illuminated area that appears white or highly transmitting
\[ \text{}Chroma = \frac{\text{Colorfulness}}{\text{Brightness(white)}} \]
Color Order Systems
Munsell Color system
Composition:
- Hue
- 10 hues (each divided into 10 subhues)
- Lightness (called Value)
- 11 steps (0: ideal black, 10: ideal white)
- Chroma
- Range depending on the hue
Eyes are all differently sensitive
Notation: H V/C
- H = Hue, V = Value, C = Chroma
- e.g. 5Y 7/12 or 5R 1/4
Munsell Color Tree from Pantone
How to communicate color ?
Pantone and other organizarions
Pantone Color-Naming System
- Used in the printing/manufactuing industry
- Swatches are used to specify colors
- Printed using 14 inks
- Useful for specifying communicating color
- Patented ! Need a license to use the list
Describing Relative Color
Color Mixing Systems
Amounts give a specification, not the resulting color
- RGB value \(\{100, 20,90\}\) in your screen \(\neq\) RGB \(\{100, 20,90\}\) in my screen
- CMY value \(\{90, 10,50\}\) in your printer \(\neq\) CMY \(\{90, 10,50\}\) on my printer
- RGB value \(\{100, 20,90\}\) in my screen \(\neq\) CMY \(\{90, 10,50\}\) on my printer
HSV/HSL Spaces
- Hue: color
- Saturation: measure of chroma
- Value or Lightness: measure of lightness
HSV/HSL difference:
- HSL: maximum lightness = white
- HSV: maximum Value = "intense" color
HSV \(\leftrightarrow\) RGB
YCbCr Space
- Y is Luminance \(\simeq\) Brightness
- Cb is related to blue Chrominance
- Cr is related to red Chrominance
Used a lot in color compression
Decomposition
Divide image in RGB
Divide Image in HSV
Divide image in YCbCr
Use Case - Color Segmentation
HSV or YCbCr can be used in Image and Video Processing (e.g., skin segmentation)
However, as RGB and CMYK, resulting color is only relative to capture condition
Describing Standard Color
- Use an universal way (i.e., numbers) to communicate color instead of using names
- Need to standardize the color forming conditions
- Illuminant (viewing lighting)
- Observer (the human visual response to a given stimulus)
- Object reflectance (\(\lambda\)-dependent spectral measurement)
- Colorimetry
CIE Standard Observers
- Light sources (primaries): \(435.8nm\), \(546.1nm\), \(700nm\)
- CIE 1931 Standard Colorimetric Observer
- \(2^o\) visual angle (\(2^o\) Observer)
- \(17\) color normal observers
- CIE 1964 Standard Colorimetric Observer
- \(10^o\) visual angle (\(10^o\) Observer)
- 76 color normal observers
CIE Standard Iluminants
- D: Different types of daylight
- D50 (\(\color{cyan}{5003K}\)) (warm daylight)
- printing/graphic arts
- D65 (\(\color{blue}{6504K}\)) (natural daylight)
- art/film/photography
- D50 (\(\color{cyan}{5003K}\)) (warm daylight)
- A: incandescent lamp (\(\color{orange}{2856K}\))
- F: Fluorescent light
- F2 (4230K), F11 (4000K)
Color Temperature
Temperature uses as a reference an ideal object called "Black Body"
Definition: Black Body
An object that absorbs completely heat and light, and radiates the energy back.
It radiates light when heated.
CIE Standard Iluminants
- Warm: \(T \lt 3300K\)
- Dormitory, Restaurant, Hotel, Coffee Shop
- Intermediate: \(3300K\lt T\lt 5300K\)
- Stores, Shcool, Libraries
- Cold: \(T\gt 5300K\)
- Offices, Hospitals
Computing XYZ Tristimulus Values
\(XYZ\) Tistimulus Value \(\to\) amounts of 3 specified stimuli required to match a color
\[ \begin{aligned} X &= k\int_{\lambda}I(\lambda)R(\lambda)\bar x_{\lambda}d\lambda\\ Y &= k\int_{\lambda}I(\lambda)R(\lambda)\bar y_{\lambda}d\lambda\\ Z &= k\int_{\lambda}I(\lambda)R(\lambda)\bar z_{\lambda}d\lambda\\ \end{aligned} \quad k = \]
Chromaticity Diagrams: CIE 1931
All Hues are perceivable by the standard observer
MacAdam Ellipses
Where are the grey ? Brown ? No light info !
Uniform color spaces
- CIE XYZ Space is not perceptually uniform
- equal perceptual differences between colors \(\neq\) equal distances in the XYZ space
- CIE recommenced uniform color spaces
- CIE 1976 \(L\times u\times v\)
- CIE 1976 \(L\times a\times b\)
Example
In matplotlib:
CIE \(L\times a \times b\) Space
\[ L^* = 116f(\frac{Y}{Y_n})-16 \begin{cases} a^* = 500[f(\frac{X}{X_n})-f(\frac{Y}{Y_n})]\\ b^* = 200[f(\frac{X}{X_n})-f(\frac{Y}{Y_n})] \end{cases} \]
\(L\times C\times h\)
What's left ?
- Modeling cognitive effects or phenomena
- How to obtain absolute color attributes (brightness and colorfulness) ?
- Debate about accuracy of \(\bar x, \bar y, \bar z\) Color Matching Functionc
- Representativeness of the population used for the experiments
- Limitaations of equipment used for the experiments
- How to understand color perception of "color anomalous" observers ?
