PA 2 Flashcards
1 millimeter = 1 000 000 nanometers
Whole spectrum of visible colours covers …
400nm
The wavelength of the light reflected only determines the hue which is seen.
Perceived colour is also determined by the:
- intensity of reflected light (brightness)
- saturation of the colour (how much white light is mixed in with pure hue
wavelenght of light determines what
the hue (colour)
opponent process theory in colour
3 opponent processes
- Red - Green
- Blue - Yellow
- Black - White
How are the trichromatic and opponent processes theories combined into one theory
Dual process theory
the retina / photoreceptors process info trichromatically
The Ganglion cells / LGN process info as opponent process stage
Reflectance =
proportion of light reflected from the surface (% of photons reflected)
Luminance
ammount of light reflected from a surface
(illumination * Reflectance)
Illumination
ammount of light emitted from a light source
Reflectance
proportion of light reflected from the surface
Lightness constancy
The perception of constant lightness despite a change in viewing conditions
Adaptation theory of lightness constancy
1 problem
Adaptation
- Visual system becomes more/less sensitive in dull/bright conditions.
BUT
these adaptations are slow, can’t account for fast changes in lighting
Unconscious inference theory of lightness constancy
1 problem
- Prior experience allows us to estimate illumination
- BUT
we are not sensitive to absolute levels of illumination
Relational theories of lightness constancy
Luminance rations determine lightness perception
Evidence for Relational theories of lightness constancy
- Participants matched luminance ratios rather than absolute luminance
- Edges important e.g. Craik-Cornsweet-O’Brien illusion
Relational theories
Retinex theory (Land & McCann, 1971)
- Luminance ratios are calculated at edges
- Gradual changes of luminance are ignored (assumed to be due to changes in illumination)
Problems for relational theories
-1. Scaling problem. a ratio of 5:1 of two greys to grey/black could have infinite combinations (Grey and white or black and grey)
2. Illumination edges: e.g. Retinex theory assumes all changes in illumination are gradual – can’t account for sudden changes in illumination (e.g. when a change in a light source is abrupt, we should interpret this as an edge in Retinex theory)
How can scaling problem be solved
solved via anchoring heuristic
Assume highest luminance is white, then scale all other regions relative to this (white balance)
Illumination vs reflectance edges
Reflectance edges = neighbouring regions have different reflectance – e.g. different material or paint
Illumination edge = – neighbouring regions receive different amounts of light e.g. shadows, spot lights, change in orientation
2 heuristics to explain Illumination vs reflectance edges
- Fuzziness heuristic = Illumination edges are often fuzzy, with reflectance edges being sharper
- Planarity heuristic = If depth information indicates that two regions are not coplanar (i.e. don’t have the same 3D orientation), likely to be an illumination edge
- Ratio magnitude heuristic = If luminance ratio is very high, likely to be an illumination edge
Colour constancy
Colour constancy = perception of constant surface colour despite changes in illumination and viewing conditions
Illumination spectrum =
amount of incident illumination at each wavelength
Reflectance spectrum
= proportion of light reflected at each wavelength
Luminance spectrum
amount of light reflected at each wavelength
How did land study The effect of surroundings on colour constancy
Conclusion
Land (1983) presented Mondrian patterns illuminated by different coloured lights
When whole pattern visible - single patch not perceived to change colour as the illuminant changed
When surroundings masked – single patch was perceived to change colour as the illuminant changed
Surroundings play important role in colour constancy as they help us to calculate the illumination spectrum and correct for it.
How this is done is not well understood
