PSYU2247 - Colour Vision week 7 PDF

Title PSYU2247 - Colour Vision week 7
Author Timothy Courtenay
Course Perception
Institution Macquarie University
Pages 8
File Size 213.8 KB
File Type PDF
Total Downloads 17
Total Views 157

Summary

Download PSYU2247 - Colour Vision week 7 PDF


Description

Colour Vision Wavelength and Colour What is wavelength? –

Light is (kind of) a wave



Wavelength is the distance between the crests (nm)

What is intensity? –

Intensity is the wave’s “height”



Approximates to brightness

Different pure, single wavelengths appear to us as different colours –



Wavelength is an objective property of the stimulus •

Short wavelengths look purpleish/blueish;



Medium wavelengths look greenish



Long wavelengths look reddish

Colour is a subjective property of the percept

Colour misconceptions •

“There are 7 colours: Red, Orange, Yellow, Green, Blue, Indigo & Violet” –



Newton did divide the spectrum into 7 different named categories, but in reality the spectrum is a continuum

“White and black aren't colours: they’re shades” –

Colour is an experience of a combination of wavelengths of light



Therefore, black and white qualify

Reflection of light  

Objects reflect light, or absorb (subtract) it They appear to be= - …The colour of the light that they reflect - …black if they absorb lots of light at all wavelengths, and reflect little - …white if they absorb little light, and reflect lots at all wavelengths - …yellow if they absorb all else

Additive Colour Mixtures   

We can add various pure wavelengths of light together to form new colour Although the combination of short & long wavelength (blue & red) light makes purple, the wavelengths of light have not changed Light containing all wavelengths together appears white

Subtractive colour mixtures •

Mixing paint is a subtractive process



Most paint reflects more than just a single, pure wavelength of light



Blue paint reflects mainly short wavelengths (blue) of light, but also some of the neighbouring wavelengths (purple & green)



Yellow paint reflects mostly yellow light, but also a bit of green & orange



Putting both together, the new patch reflects only green



The more colours of paint you add, the more light is absorbed/subtracted and the darker the patch looks



Rarely in nature is there one single pure wavelength

Photoreceptors •



Rods -

Only one type None in central fovea High sensitivity Night vision Scotopic

-

Several types, each tuned to a different wavelength Mostly in fovea Lower sensitivity Day vision Photopic

Cones

Scotopic (night time) Colour Vision



At low luminance (e.g. night vision), cones are not sensitive enough to respond - rods only



Our colour vision with rods alone is poor: almost non-existent



“principle of Univariance”

Colour Vision with One Receptor Type •





Receptors show broad wavelength tuning –

For rods, probability of absorbing (“catching”) photons is highest for wavelength of ~500nm (greenish), smaller probability of absorption for other wavelengths



Hence, peak response at ~500nm, but smaller response to most wavelengths of visible light

Principle of Univariance –

A receptor’s activity is related to the number of photons it catches, not to the type (wavelength) of photon



Same response to a single wavelength at 450 (blueish) or to 650nm (orangeish) light at 160cd/m2



Response to 550 nm (green) light can be the same also if intensity is reduced to 80cd/m2

One lone receptor can only tell light from dark –

Any wavelength can look like any other wavelength, given the right intensity



Hence bad colour vision at night (rods only)

Photopic (day time) Colour Vision •

Rods are all “bleached out” (over-stimulated): It’s down to the cones



Unlike rods, there are several different types of cone, each tuned to a different wavelength



Colour vision in daylight is good: humans can reliably discriminate at least 200 wavelengths



How many different types of cone are there? 200?

Young’s Colour Theory

The Young-Helmholtz theory states that within your eye are tiny cells that can receive waves of light and translate them into one of three colors: blue, green, and red. These three colors can then be combined to create the entire visible spectrum of light as we see it.

A Two receptor system •

What if we had 2 cone types? –

Colour represented by the relative activation of the 2 cone channels (call them “M” and “L”)



As wavelength changes, pattern of relative activation changes





Short wavelengths would produce more activity in M cones than in L cones



Medium wavelengths would produce equal activity in each cone type (see fig.)



Long wavelengths: L activity > M activity

As intensity changes, activity changes equally in both channels: hence the relative activation remains the same •

Bright medium wavelengths produce equal high activity in each cone type



Dim medium wavelengths produce equal low activity in each cone type



Allows us to tell a wavelength change from an intensity change



Two different physical stimulus configurations that appear identical



Physics is different but perception is identical



For a 1- receptor system (e.g. rods at night) all wavelengths can be metamers for each other (so long as you get the intensities right)

Metamers

With a 2-cone system… – …we could match any given wavelength by adjusting the intensities of almost any two other wavelengths (metamers) – …there would be a single wavelength that appears grey. This is known as a “neutral point”, where the firing of both channels is equal Colour matching



Colour Matching Experiments (Young, 1802) –

To match any single, pure wavelength of light (i.e. create a metamer), humans need 3 (or fewer) “primaries” of adjustable intensity



Any 3 wavelengths of light can be used (so long as they are suitably different)





Like your TV screen, or the projector in this room



Most use single wavelengths that appear Red, Green & Blue

From this we infer that most observers have 3 cone types

Young-Helmholtz Trichromatic Theory of Colour Vision •

Developed many decades before neurophysiologists discovered & measured the 3 separate cone types



There are only 3 receptors, each broadly tuned to wavelength



Perceived color depends on the relative strength of their activation

Three Cone Types •

Long (peak 560 - reddish)



Med. (peak 530 - greenish)



Short (peak 420 - blueish) – Blue light is often blurred •

Few in the fovea



More sparse

Colour Misconceptions #3 •

“Red, Yellow and Blue are the primary colours”



“Red, Green and Blue are the primary colours” •

If we’re talking about the 3 different photoreceptor types, we should really refer to Short (~420nm), Medium (~530nm) & Long (~560nm) wavelengthtuned photoreceptor types.

Trichromatic Metamers



Colour encoded by the pattern of relative activation of each channel



Like population coding in spatial vision, olfactory perception, etc.



Observe the levels of cone activation for 600nm (red), 505nm (green) and 545nm (yellow)



Imagine what the cone activation levels would be like if red (600) and green (505) wavelengths were presented together –



Add together the Short wavelength cone response to red & to green, then do the same for the other two cone types

Activation pattern for red + green presented together is the same as for yellow presented alone

Colour Deficiency •

Different types associated with different cone types



Anomalous Trichromats –

Have 3 types of cone, but responses are different from most (normal) people, due to deficiency in one cone type •



Dichromats –

Lack one cone type entirely •



Protanomaly (red/long cone), Deuteranomaly (green/medium cone), Tritanomaly (blue/short cone)

Protanopia/Protanopes (red/long cone), Deuteranopia/Deuteranopes (green/medium cone), Tritanopia/Tritanopes (blue/short cone)

But some people are truly colour blind (i.e. can only tell light from dark) –



Monochromats •

Cone monochromats have only one type of cone & lack colour vision day & night



Rod monochromats have no cones at all Highly sensitive - need dark glasses in daylight

Cerebral Achromatopsia •

Colour Misconceptions

Damage to area V4 can cause complete “colour blindness”



Colour blind people see only black and white”



Depends what you mean by “colour blind” –

The majority of colour deficient people (dichromats and anomalous trichromats) can discriminate many colours



Monochromats (only one cone, or only rods) and those with complete Cerebral Achromatopsia can only discriminate light from dark (truly “colour blind”)

Colour Opponency (Hering, 1872) •

Observations –

When asked to select “pure” colours, people pick 4, not 3



Humans describe the world using 4 colours •



Reddish-blue, or yellow-green, but never reddish-green or blueish-yellow

Inference –

Two opponent colour axes: red-green and blue-yellow



For full colour space, we need to include the black-white axis



Grey is neutral: neither red nor green; neither blue nor yellow; neither black nor white

Dual Process Theory •

Hurvich & Jameson (1957): dual process theory wherein trichromatic receptors (cones) feed an opponent process – Luminance (B&W) Green

=

Long + Mid

or

Red +

– Red&Green Green

=

Long ÷ Mid

or

Red ÷

– Yellow&Blue Blue

=

(Long + Mid) ÷ Short or

(Red + Green) ÷

Physiology of Opponent Processing



Cells have been found in LGN and V1 which show antagonistic cone inputs



Some also show a centre-surround arrangement - “double opponent” cells



R/G cells are Parvocellular, B/Y are Koniocellular, B/W (luminance) are Magnocellular



V1 hypercolumns contain “blobs”: cylinders where opponent colour cells are located



Blobs - class of neurons with very simple circular-surround fields; projections from parvocellular LGN that retain color information

Colour constancy •

Objects generally look the same colour in a wide variety of lighting conditions



Land (1977) – Isolated coloured patch appears to change colour when lit by different wavelengths – When this patch is part of a Mondrian Pattern, it appears to be the same colour regardless of the illumination...


Similar Free PDFs