9 Lecture 10 Color Vision Not All Animals Have

9 lecture 10 color vision not all animals have color vision. mammals tend to have limited color vision. some that do: primates

9
Lecture 10
Color vision
not all animals have color vision.
mammals tend to have limited color vision. Some that do:
primates
cats
squirrels
prairie dogs
Also, many species of:
birds
fish
amphibia
reptiles
arthropods
have quite highly developed color vision.
Color perception is based on the wavelength of light.
The dimensions of color
Perceptual dimension/sensation Physical quantity
Hue wavelength
Brightness intensity
Saturation spectral purity
light can contain mixture of wavelengths
if pure = monochromatic = fully saturated
We're going to explore some phenomena of color vision, then discuss
some theories to explain these phenomena, and relate the phenomena to
the underlying neural basis of color vision.
Mixing colors:
First, what happens when lights of 2 wavelengths combined (light
entering eye = 2 wavelengths)
what do you see? => only one color.
There are two ways to do this. Both affect the light entering the eye.
Additive = add two lights of different wavelengths together.
Subtractive we'll get to in a moment.
A. Additive color mixtures
What color do you see when two lights of different wavelengths are
added together? There are several ways to predict what that will be.
1. Color circle (fig 5.3, see also 5.2) (Sekuler 6.7)(Lect8.ppt)
First invented by Newton.
wavelength ordered around circumference
radius = degree of saturation
complementary color => if mixed => gray (or white)
mixing noncomplementary colors => percept of intermediate wavelength
NOTE: some colors, namely purple, have no pure wavelength!! lies
within "gap",
But other colors can be the product either of pure wavelength or of
mixture of other wavelengths.
= metamers
Metamer = pairs of lights of physically different wavelengths that
have the same effect on the visual system and hence appear identical
in color
e.g. mix red & green light => appears yellow
yellow and red/green are metamers, or metameric pairs
match = metameric match
If colors are matched in unequal amounts, => weighted average (draw
line connecting two points on color circle, then point on that line
corresponds to the mix) Fig 5.3
2. Trichromatic approach.
Color circle = one way to describe this, predict what metamers will
be, etc.
An alternative: the trichromatic approach to color vision.
= Any color can be produced by combining various amounts of 3
specially chosen colored lights
= primary colors
= blue, green, red, usually - though other sets are possible
Examples of additive color mixing:
TV -
Color tv's have only 3 colors -- closely packed pixels of red, green,
blue -- check it out, use a magnifying glass
Art -
pointillism
e.g. Seurat, Signac - tiny dots of paint, fuse at a distance
B. Subtractive color mixing
Additive = directly mixing the lights. A good way to understand what's
going on, but in fact, most of the colors we see in the world are the
result of differential absorption & reflection of incident light based
on it's wavelength. In other words, what does blue paint do? It
absorbs all wavelengths except blue, which it reflects.
Subtractive color mixing is what happens when you mix pigments
together. You effectively "subtract" all the wavelengths absorbed by
the second pigment from those reflected by the first. All wavelengths
except those reflected in common by both pigments will be absorbed.
Subtractive color mixtures are hard to predict because you need to
know a lot about exactly what light is being reflected by each
pigment, so that you can determine the overlap between the two
pigments.
Afterimages:
Effects of visual stimuli persist after its physical termination
A particular kind of aftereffect = an afterimage, where you actually
see something that isn’t there.
2 kinds:
positive:
same light-dark & color pattern as actual image
e.g. afterimage of the flash from a flashbulb
negative:
the pattern is reversed, like film negatives
Two demos:
black & white
color
If color:
"negative" afterimage = complementary color
so, stare at red => green afterimage, etc.
(demo)
What's happening?
region of the retina(? could also be a higher visual area) that is
stimulated becomes adapted, fatigued, or less sensitive to the
particular color. this is functionally equivalent to making the same
retinal area more sensitive to its complementary color.
Simultaneous contrast:
Colored patch on gray background appears tinged with complementary
color
continued fixation of a colored stimulus patch that appears against a
neutral or gray background, the edge of the background appears tinged
with the complementary color.
Why?
Related to negative afterimages.
Adaptation of one color => increased tendency to see complementary
color. Don't see this in the center because the actual color is still
so strong. But around edges, on neutral background, complement begins
to appear. Eye movements (fixation jitter) contributes to this.
color patch loses "saturation", complementary is induced, and since
eyes move slightly, even during fixation, the retinal image of the
patch is not perfectly stationary. So complementary color begins to
appear around the edges.
Demo.
Color constancy:
Light sources have different spectral compositions:
sun - broad spectrum
tungsten lightbulb - slightly yellow
fluorescent - slightly blue
=> the spectral content of light reflected off of objects varies under
these conditions
Nevertheless, within broad limits, the perceived color of objects =
constant
= Color constancy
= tendency for an object's perceived color to remain constant, despite
changes in the wavelength of the illuminating light
How does this happen?
Don't know for sure, but know some factors that contribute:
1. Effects of background
visual system seems to be able to take into account, and correct for,
spectral changes that affect the entire visual scene equally.
In other words, your brain can tell something about the spectral
content of the illuminating light by looking at the whole scene. This
is then factored in and compensated for.
This is useful, because under natural conditions, the illumination of
the scene changes, as the sun rises & sets, for example.
2. Color adaptation
over time, adapt to dominant wavelengths
become less sensitive to tungsten's yellow light, etc.
What is color constancy good for?
Color constancy is the first of a series of different types of
constancies we'll encounter.
= > strongly influences our perception of a stable environment
helps us recognize objects upon repeated presentation under differing
conditions
Okay, now that we've talked about a few of the basic phenomena of
color vision, we're ready to turn to the implications of these
results:
Theories of color perception:
1. Trichromatic receptor theory / Young-Helmholtz Theory
This was first proposed by Thomas Young, an English scientist, in 1802
(Young also deciphered the Rosetta stone)
Revived in 1866 by Herman von Helmholtz
Noticing the phenomena of color mixture, Young & Helmholtz thought
that it suggested certain structural, functional, and neural
mechanisms of the retina.
lights of 3 distinct wavelengths = sufficient to produce the complete
visible spectrum
=> implication = 3 corresponding types of receptors in the human
retina, with different spectral sensitivity
= trichromatic receptor theory
Extended Young's theory to postulate that those three receptors =
tuned for blue, green and red.
graded responses to those wavelengths, overlapping sensitivities
We've already talked about the cone cells in the retina, of which
there are 3 types. But, this wasn't discovered until many years later.
Cones:
S, M, L cones
3 different pigments
absorption curves: Fig. 5.9 Lect8.ppt
note that curves overlap
trichromatic theory says these 3 cone types account for many color
phenomena (a Linking Hypothesis)
- afterimages
= fatigue one cone type
=> mess up ratio
=> complementary afterimage
- mixing
mixture of wavelengths => same net pattern of activation of 3 cone
types as some other wavelength
e.g.
suppose there is some unique wavelength that activates all 3 cone
types to different degrees,
say,
s=25% of maximum
m=50% of maximum
l=10% of maximum
this same effect can be achieved with various amounts of 3 different
wavelengths,
s-25%, m 50%, l 10%
Lect 9 Color vision continued
Theories of color perception:
1. Trichromatic receptor theory / Young-Helmholtz Theory
This was first proposed by Thomas Young, an English scientist, in 1802
(Young also deciphered the Rosetta stone)
Revived in 1866 by Herman von Helmholtz
Noticing the phenomena of color mixture, Young & Helmholtz thought
that it suggested certain structural, functional, and neural
mechanisms of the retina.
lights of 3 distinct wavelengths = sufficient to produce the complete
visible spectrum
=> implication = 3 corresponding types of receptors in the human
retina, with different spectral sensitivity
= trichromatic receptor theory
Extended Young's theory to postulate that those three receptors =
tuned for blue, green and red.
graded responses to those wavelengths, overlapping sensitivities
We've already talked about the cone cells in the retina, of which
there are 3 types. But, this wasn't discovered until many years later.
Cones:
S, M, L cones
3 different pigments
absorption curves: Fig. 5.9 ([lect8.ppt])
note that curves overlap
trichromatic theory says these 3 cone types account for many color
phenomena
- afterimages
= fatigue one cone type
=> mess up ratio
=> complementary afterimage
- mixing
mixture of wavelengths => same net pattern of activation of 3 cone
types as some other wavelength
e.g.
suppose there is some unique wavelength that activates all 3 cone
types to different degrees,
say,
s=25% of maximum
m=50% of maximum
l=10% of maximum
this same effect can be achieved with various amounts of 3 different
wavelengths,
s-25%, m 50%, l 10%
Theory 2: "opponent process theory"
Ewald Hering 1920
also says 3 independent mechanisms, but = 3 pairs of opponent
processes
blue - yellow
green - red
white - black
Each "channel" = capable of signaling either of its two values (e.g.
green or red), but not both simultaneously
Evidence:
afterimages
-channel = receptor pair
fatigue one, get exaggeration of the other
SYNTHESIS:
Hurvich & Jameson - 1955
3 types of retinal receptors
-> connected in opponent fashion
blue-yellow
green-red
white-black (achromatic)
=> antagonistic interactions
=> 2 Stage Process
1. receptors
2. neural opponent circuitry
Physiological basis:
What is mechanism for the opponency that is apparent in various
aspects of color perception such as afterimages?
Retina
Cones:
photopigment = retinal + cone opsin
the opsins in each cone type are different
We already mentioned that
X (P) ganglion cells = color sensitive
LGN (that receive input from X) = color sensitive
So, what are the response properties of these neurons vis-a-vis color?
=> They are color opponent!
get strongest excitation, inhibition, depending on the wavelength:
(draw response as function of wavelength)
1. cone cell

2. RGC –X cells (P cells)
8 types:
2 pairs of color opponencies
R/G R+/G-, G+/R-
B/Y B+/Y-, Y+/B-
on center & off center

Specific circuitry for physiological basis of 2 stage theory: - 5.11
This kind of receptive field = single opponent, because there’s only 1
color comparison, between center & surround.
How would you wire these up? (Remember, cones, like rods, are
inhibited by light)
=============================================================




(and, of course, have input come from different locations to get the
center-surround organization).
Cortex:
Some neurons not tuned for orientation, shape, movement
= exclusively color
have double oppponent rf's
center & surround + color opponency within each part e.g.
center: R+, G-
surround: G+, R-
(Note to JMG: blob/interblob color segmentation may not be true)
Defective Color Vision
3 cone pigments <-> 3 genes
deficiencies = genetic, inherited
Most common problem involves Red & green pigments
genes for both = X chromosome
= sex linked recessive
affects 9% of males, 0.5% females
3 major classes of color vision problems
- anomalous trichromatism
- dichromatism
- monochromatism
Anomalous trichromatism
- individuals use color mixing combos of primary colors that differ
from the norm
2 forms:
protoanomaly
deficiency in L pigment
=> reduced sensitivity to red
deuteranomaly
likewise for M pigment & green
need more red or green, respectively, in color mixing tasks
Dichromatism
perform color mixing tasks using only 2 colors, not the normal 3
2 types
protanopes
lack L pigment
= insensitive to red
deuteranopes
lack M pigment
= insensitive to green
both types see blue & yellow okay, and both confuse red and green
What do deuteranopes & protanopes see?
One unusual subject was a deuteranope in one eye and normal in the
other
color mixing task -
view mixture with color blind eye, then, with normal eye, adjust
second stimulus to match its color
=> Fig 5.13
=> wide range of wavelengths look yellow
3rd, rare kind of dichromat =
tritanopia
= lack of S cone
deficiency in distinguishing blues & yellows & grays
Monochromatism
extremely rare
= true color blindness
can't distinguish wavelength at all
=> probably have abnormal number & assortment of cones
tend to have other visual abnormalities as well
What do these defects say about our theories of color vision?
Consistent with trichromatic theory, specific color defects arise from
the relative lack of particular cone pigments
Consistent with post-receptor opponency, defectiveness at seeing red
=> defectiveness at seeing green, etc.