1/207
Looks like no tags are added yet.
Visible spectrum
Electromagnetic energy between 400 and 700 nm
Monochromatic light
Light containing only a single wavelength
Spectral colors
Colors associated with one monochromatic wavelength
Heterochromatic light
Light containing many different wavelengths
What does a prism do?
Disperses white light into its component colors
Spectral reflectance
The ratio of light reflected by an object at each wavelength (what we 'see')
Does the visual system respond to the total amount of light reflected or the percentage reflected?
The percentage reflected
Why do surfaces keep their color when everything gets dimmer?
The RATIOS of reflected light between surfaces stay constant (part of color constancy)
Fluorescent vs. incandescent light
Fluorescent peaks strongly in blue and green; incandescent rises gradually and peaks in long (red) wavelengths
Is color an inherent property of an object?
No; it's the brain's interpretation of the wavelengths reaching the retina
Hue
The color quality of light (red, green, blue): 'what kind?' A quality, not an amount
Saturation
The purity of light: 'how much hue?' An amount, not a quality; higher saturation = less white light mixed in
Brightness
The perceived amount or intensity of light present
Lightness
The psychological experience of light reflected by a surface
Achromatic surface
Reflects all wavelengths equally; no color (no dominant hue or saturation)
Reflectance of an achromatic surface seen as WHITE
90% or higher
Reflectance of an achromatic surface seen as GRAY
About 50%
Reflectance of an achromatic surface seen as BLACK
10% or lower
Color circle
Perceptual map: hue on the perimeter, saturation increases outward, neutral gray at the center
Order of colors from shortest to longest wavelength
Blue - Green - Red
Which wavelengths are most associated with 'red'?
Long wavelengths
Cone names used instead of blue, green, red
S (short), M (medium), L (long)
How is an object's color determined by cones?
The joint pattern of response across S, M, and L cones to the object's reflected wavelengths
Additive color mixing
Creating colors by ADDING wavelengths; R + G + B = white (e.g., laser light show)
Subtractive color mixing
Creating colors by REMOVING (absorbing) wavelengths; mixing paints, dyes, pigments; C + Y + M = near black
Mixing paints together is an example of…
Subtractive color mixing
Why can different combinations of wavelengths look like the same color?
Different ratios of wavelengths can produce colors that look the same, because color collapses to just 3 cone values
Principle of univariance
Many combinations of wavelength and intensity elicit the same response; one receptor can only signal the amount of light absorbed
Why does univariance require at least two cone types for color vision?
A single receptor loses wavelength information, so color needs a comparison between receptor types
What does univariance explain about scotopic (night) vision?
Why it lacks color: only one receptor type (rods) is active
Trichromatic theory (Young-Helmholtz)
Color is determined by the RATIO of activation of three cone systems
What does trichromatic theory explain?
Processing at the photoreceptor level and color blindness as missing cone types
Opponent-process theory (Hering)
Color arises from three opponent mechanisms: red-green, blue-yellow, black-white
What does opponent-process theory explain?
Why we don't see 'reddish greens' or 'bluish yellows,' why unique hues feel basic, and afterimages
Afterimages
Images seen after a visual stimulus is removed, appearing in the complementary color (opposite side of the color wheel)
How do trichromatic and opponent-process theories fit together?
Stage 1: cones are trichromatic. Stage 2: ganglion cells and LGN recombine signals into opponent channels
Red-green opponent channel: computation and pathway
L - M; parvocellular
Blue-yellow opponent channel: computation and pathway
S - (L + M); koniocellular
Light-dark opponent channel: computation and pathway
L + M; magnocellular
Cone-opponent cells (LGN)
Excited by input from one cone type in the center and inhibited by another cone type in the surround
Where are cone-opponent cells found?
LGN
Color-opponent cells (V1)
Excited by one color in the center and inhibited by its opponent color in the surround
Where are color-opponent cells found?
V1
Double-opponent cells
Cells that enhance color divisions at edges
Color vision in newborns
Limited chromatic vision that improves rapidly (nature + nurture)
Color vision and aging
Lens 'yellows,' reducing transmission of short wavelengths (less blue sensitivity); dark adaptation takes longer
Are most people with color deficiencies 'color-blind'?
No; a range of variation is common
Color deficiencies are often linked to…
The X chromosome (genetic)
Dichromacy
Two working cone systems; see color but can't make some discriminations trichromats can
Most common color deficiency
Red/green
Monochromacy
Two or all three cone types absent or nonfunctional; often with other problems like brightness sensitivity and reduced acuity
Color constancy
Inferring stable surface colors from a noisy, changing light signal (a red apple looks red in sunlight or a dim room)
How does the brain achieve color constancy?
Discounting the illuminant, ratios of reflected light, context and surrounds, top-down knowledge, chromatic adaptation in the retina
The depth problem
Everything along a cone of sight projects the same-size image on the flat retina, so the brain must infer size and distance
Cue approach
Depth is inferred by combining cues: aiding factors in a nonconscious, automatic decision
Three main categories of depth cues
Monocular, oculomotor, and binocular
Monocular cues
Depth cues needing only one eye; include pictorial and motion-based cues
Pictorial cues
Monocular depth cues from static (nonmoving) images
List the pictorial depth cues
Occlusion, relative height, relative size, familiar size, linear perspective, texture gradient, atmospheric perspective, shadowing
Occlusion (interposition)
A nearer object partially blocks a farther one
Linear perspective
Parallel lines (like railroad tracks) appear to converge toward the horizon
Motion-based monocular cues
Motion parallax, accretion, deletion, optic flow
Motion parallax
Objects at different depths move at different speeds relative to each other (e.g., looking out a side car window)
Accretion
Gradual REAPPEARANCE of a moving object after being blocked
Deletion
Gradual OCCLUSION of an object behind another as it (or the observer) moves
Optic flow
Pattern of image motion as we move forward or backward (e.g., through a car windshield)
Oculomotor cues
Depth cues from eye-muscle feedback to the brain as the eyes adjust near to far; contraction relates to distance; automatic
Accommodation
Oculomotor cue: muscles change the lens shape to focus; not precise (narrow pupil by day, poor acuity at night)
Vergence
Oculomotor cue: eyes rotate inward (converge) for near objects, outward (diverge) for far
How far is vergence a precise depth cue?
Up to about 2 meters; no vergence after 4 meters
Binocular disparity
Depth perception from the slightly different left- and right-eye views
Stereopsis
Sense of depth from the visual system comparing the two eyes' disparate images
Where is binocular vision possible?
Only where the visual fields of both eyes overlap
How do 3D movies with anaglyph glasses work?
Each lens filters out one of two overlapping images, so each eye gets a slightly different image (stereoscopic vision)
Corresponding points
A point on the left retina and a point on the right retina that match in location
Noncorresponding points
Points on the two retinae in slightly different locations
Diplopia
Double vision when the mismatch between the eyes' images is too large to fuse
Horopter
Curved region in space where objects fall on corresponding retinal points and appear single (fused)
Disparity on the horopter
Zero disparity
Crossed vs. uncrossed disparity
Crossed (+) = in front of the horopter; uncrossed (-) = behind it; both are noncorresponding points
A car image is 1 mm left of the fovea in the left eye and 2 mm left in the right eye. What kind of points?
Noncorresponding points
Correspondence problem (depth)
How the visual system knows which image in the left eye matches which in the right eye
Stereogram
One image to the left eye and one to the right, similar enough to combine into an illusion of depth
Random dot stereogram
Research tool with no recognizable object until the images are fused; shows depth processing precedes object recognition
Interocular velocity differences
An approaching object moves different amounts on each retina depending on its direction
Where are binocular disparity-sensitive cells found?
V1 (striate) and V4 (extrastriate)
Disparity-tuned neurons
Binocular cells with a receptive field for each eye that fire for noncorresponding points; tuned to crossed or uncrossed disparity
Size-distance invariance
The relationship between perceived size and perceived distance
Visual angle
The angle an object subtends at the eye; farther = smaller angle and retinal image, closer = larger
Size constancy
Perceiving an object as a fixed size despite changes in visual angle with distance
Muller-Lyer illusion
Lines look different lengths because angled fins signal an outside corner (closer) vs. inside corner (farther); misapplied size constancy
Ames room
Trapezoidal walls, floor, and ceiling make two people in opposite corners look drastically different in size
Motion
A relative change in position over time
Three dimensions of motion
Ahead, side, above/below
Minimum motion for detection (absolute threshold)
At least 1 minute of 1 degree across the retina
Are motion thresholds fixed?
No; they're context-dependent, influenced by brightness and size, and differ between periphery and fovea
Real motion
Continual change in an object's position relative to some frame of reference
Apparent motion
The appearance of motion with no corresponding real motion; a mismatch between senses and reality (illusion)
Beta motion
Sequential images perceived as realistic movement; basis of movies, flip books, scrolling LED displays
Phi motion
You perceive motion but can see the elements don't actually move; e.g., flashing perimeter lights