Color Vision - PowerPoint PPT Presentation

1 / 39
About This Presentation
Title:

Color Vision

Description:

Two samples showing the distribution of L, M, and S cones (colored as R, G, and ... midget ganglion cells (receiving input from midget bipolars whose RF centers ... – PowerPoint PPT presentation

Number of Views:336
Avg rating:3.0/5.0
Slides: 40
Provided by: Optom
Category:
Tags: color | midget | vision

less

Transcript and Presenter's Notes

Title: Color Vision


1
Color Vision Mechanisms and Perception
2
(No Transcript)
3
Parafoveal cones and rods (macaque)
Foveal cones
No rods or S-cones in central fovea
4
Two samples showing the distribution of L, M, and
S cones (colored as R, G, and B) in these
pictures in the perifoveal retina (Roorda, UH
Optometry) Relative numbers of three cones
types S-cones (5-10, but almost zero in the
central fovea) M-cones (about 30) L-cones (about
60)
5
Stage 1 Photoreceptors 3 types of cone
L
The differences between pigment absorption
curves and in-eye sensitivity results from
pre-receptor filtering (primarily absorption of
short wavelengths)
M
S
L
M
S
Pigment absorption
Each cone contains a different photopigment. Once
referred to as erytherolabe, chlorolabe and
cyanolabe, but now more often just L, M, and S
cone pigments. The opsin molecule differs in
each cone
Cone sensitivity
L
M
S
6
Scematic diagram of cone opsin protein strand
which is imbedded in and traverses the disc
membrane within the cone outer segment. Each
circle indicates an amino acid.
The red amino acids (helix 6) are the two that
determine if the photopigment will be L or M.
Changing these two produces a 16 - 24 nm shift in
peak absorption.
Yellow amino acids dimorphism produces small
changes in spectral tuning of L or M pigments of
color normals and anomalous trichromats. These
are the spectral tuning sites.
7
The Opponent Color Theory
The Young-Helmholtz trichromatic theory provides
a basis for all three-dimensional color
quantifications systems. Since the
multidimensional physical spectra of stimuli are
immediately coded by just 3 cones types, there
can only be three independent variables that
determine all color (response of the L, M and S
cones). Color matching, metameric matching is
explained by trichromacy. There are, however,
many color phenomena that cannot be explained by
the trichromatic theory. Hering, early 20th
century (1920) argued for a different color
theory called the color opponent theory. He
argued that there were 4 elemental colors (R,Y,
G, and B) not three. He also noted the pairing
of R G, and of B Y. 1. There is no color
that appears to a mixture of RG or of BY. 2.
After prolonged exposure to R, the after-image is
always G, and vice versa. A similar pairing of
colored after-mages occurs for B Y. 3. It is
possible to induce a color by surrounding an
achromatic color by some other color. Red
induces Green and vice versa and Blue induces
Yellow and vice versa.
8
Colored After-images
9
red
black
blue
green
white
yellow
10
(No Transcript)
11
(No Transcript)
12
Demonstration of brightness induction effect
perceived brightness is not simply determined by
the amount of light, but by the contrast.
Light induces dark and dark induces light.
13
Brightness induction effect
Demonstration of how color and brightness are
affected by the background.
14
Induced color (color contrast effects)
15
Color naming experiments (subjects had to
indicate what of each color appeared to be Red,
Green, Blue or Yellow). Also, they indicated
saturation of the colors.
Notice that there are no colors that appear to be
a mixture of Red and Green, and the same is true
for Blue and Yellow. Redyellow, and Green
yellow coexist, as do Green Blue and Red blue.
16
Coding color by the relative response amplitudes
of the different cone types
Cone spectral sensitivity curves
Each individual cone is completely color-blind,
and color information is coded in the relative
responses of the three cone types. The most
simple relative code is the difference between
two or more cone types, e.g. L - M, or M - L.
There is a lot of evidence to support the idea
that this is exactly how the visual system codes
color. We find post-receptoral neurons in the
retina and LGN that have opposing sign inputs
from diffferent receptor types.
S
L
M
17
Neural explanation for color opponency
Post-receptoral mechanisms opponent color theory
B
R
luminance
G
Y
A
R/G

B/Y

-
-


S cones
M cones
L cones
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
  • Schematic showing the retinal, LGN and V1
    projections of three types of retinal ganglion
    cell
  • midget ganglion cells (receiving input from
    midget bipolars whose RF centers receive input
    from individual cones project to Parvo-cellular
    layers of LGN, red-green opponent cells).
  • Parasol cells (large dendritic trees and RFs,
    project to magnocellular layers of LGN,
    achromatic cells)
  • Small-bi-stratified blue-on cells (receive input
    from two sub-layers of inner-plexiform layer, and
    project to the sparsely populated inter-laminar
    kioniocellular layers of the LGN)

22
Cross-section through V1 showing the projection
and color-specificity of Parvo-, Magno- and
Konio-cellular layers of the LGN
1
2/3
4A
4Ca
4Cb
5/6
P
M
K
23
Problems with classic opponent color theory (1)
Gives no redness at very short wavelengths and
no unique blue.
Perception Opponent Model
B
R
Response
Y
G
Wavelength
Unique yellow
Unique green
24
Simple Fix
Responses of opponentmechanisms
B
R
R/G
Response
Y
G
-


Wavelength
Unique yellow
Unique blue
580nm
Unique green
470 nm
515 nm
Add short wavelength input to the R/G opponent
mechanism (with same sign as L cone) to get (1)
redness at very short wavelengths, (2) a unique
blue.
25
Limitations of Color Opponent Model 1. Does not
explain why more saturated colors appear brighter
than equiluminous desaturated colors. Added
extra neural stage made brightness the sum of
the achromatic (A) and opponent (R/G, B/Y)
mechanisms. 2. Does not explain color
assimilation 3. Does not explain color contrast
sensitivity function. Basically, trichromatic
color theory and opponent color theory fail to
include the spatial component of color vision.
26
Color Assimilation if spatial detail is small,
clor bleed into each other
Notice that the yellow stripes appear clearly
yellow when wide, but they lose their color when
very thin.
27
Notice that red surrounded by blue looks more
blue than red surrounded by yellow.
28
Commercial application of color assimilation
Notice that mesh packaging is typically the
desired color of the fruit or vegatable in the
package.
29
(No Transcript)
30
Comparison of Color and Luminance Contrast
Sensitivity
2
CS
1. Reduced Spatial Resolution 2. No Low SF
fall-off in CS
1
20/15
20/100
Spatial Frequency
31
With increasing SF and TF, less and less color
can be seen.
The rich spectrum of visible colors that we are
used to seeing are not present for all stimuli.
If a light flickers red/green at a fast rate
(e.g. 30Hz), no color change will be seen, only a
brightness change. Likewize, if a series of
red/green stripes was made very fine (e.g. 30
c/deg), they would appear to be stripes of
differing brightness but the same color.
32
Color adaptation and color constancy
maintaining perceptual constancy as the
illuminant changes (e.g. from noon to dusk, from
fluorescent to incandescent).
Bright adapting stimulus
Post-adaptation white
White w/o adaptation
All colors within dotted area can appear white
after color adaptation.
Why is the system so adaptable? This seems to
make no sense. Surely we would want a given
stimulus spectrum to always produce the same
color sensation.
33
Away from sun
Reflected Spectrum
Towards sun
Illuminant
White paper at noon
The reflected spectrum, the one that we see, is
determined by the properties of the surface AND
the spectrum of the illuminant. Thus, if the
illuminant changed (e.g. noon to dusk), the
object spectrum will change. If there was no
color adaptation, the perceived color of objects
would change throughout the day. With color
adaptation, color constancy of objects, not
spectra, is achieved.
White paper at dusk
34
Additional Color information 1. Color labels
and their physical counterpart Hue
Wavelength Saturation Purity Brightness
Intensity 2. Although hue is dominated by
wavelength, it also changes with stimulus
luminance. Increase luminance and reds and
yellow/greens look more yellow, cyans and violets
look more blue (Bezold Bruke Effect). 3. A
saturated color will appear brighter than an
equal luminance white (Abney effect).
Explanation brightness is determined by the
activity of the achromatic AND color opponent
mechanisms and not just the achromatic channel
(which determines luminance).
35
Color discrimination How good are we?
Protanomalous deuteranomalous trichromats
Normal trichromats
1. Wavelength Discrimination
36
Purity Discrimination remember that Dp(max) 1.0
Neutral points
37
S-cones are special
None in the central fovea
Not many of them (7 of cones)
Poor spatial resolution
Slow
Susceptible to disease
Rod-free central fovea
38
Short Wavelength Cones are different
1. Slower
DEMO
2. More sensitive to disease
Isolating the S-cone response seeing with S-cones
Spectral sensitivity curves for fully dark
adapted L, M, and S cones
Cone spectral sensitivities in presence of bright
yellow light
SWAP (short wavelength perimetry) for early
detection of Glaucoma, SWS cones seem more
susceptible.
39
DEMO
Benhams top illusion Rotate quickly, thus at
any retinal location the stimulus is going on and
off, and since the S-cone response lags behind,
at any point in time, there will be an imbalance
between S and LM cones.
Cone response
yellow
blue
L,MS cones
time
stimulus
Write a Comment
User Comments (0)
About PowerShow.com