Study Guides (248,610)
Canada (121,635)
Psychology (1,882)
PSYB51H3 (55)
Midterm

PSYB51 Ch5-8 (Midterm2).docx

47 Pages
208 Views

Department
Psychology
Course Code
PSYB51H3
Professor
Matthias Niemeier

This preview shows pages 1,2,3,4. Sign up to view the full 47 pages of the document.
Description
PSYB51- Midterm 2 Notes CH. 5-8 Chapter 5 Perception of Color  Color not a physical property of things= a creation of the mind due to our particular vision systems  Humans see narrow range of the electromagnetic spectrum (400 – 700nm wavelengths) apparent color of what we see is correlated with the wavelengths of the light rays reaching the eye from that bit of the world  Most light we see= reflected light the more light absorbed by an object= the darker the object appears the color we see depends on the wavelength of the light that is reflected of the surface  3 steps of color perception 1. Detection= wavelengths must be detected 2. Discrimination= ability to tell the difference between one wavelength (or a mixture of wavelengths) and another 3. Appearance= assign perceived colors to go with lights and surfaces in the world want those perceived colors to go with the object and not change dramatically as the viewing conditions (in the sun/in the shadows) change  STEP ONE: COLOR DETECTION o 3 types of cone photoreceptors- differ in the photopigment they carry, and as a result they differ in their sensitivity to light of different wavelengths o Single photoreceptor shows different responses to light of different wavelengths but the same intensity o 3 cone types, named for location of the peak of their sensitivity on the spectrum S-cones (short, peak at 420nm/blue), M-cones (medium, peak at 535nm/green), L-cones (long, 656nm/red)  assigned colors to each cone are not really accurate, as long cones for example actually peak at a yellowish color  s-cones= relatively rare, and are less sensitive than M and L cones  combination of sensitivities of 3 types of cones= overall ability to detect wavelengths from about 400 to 700nm o cones work at daylight (photopic) light levels and one type of rod photoreceptor works in dimmer (scotopic) light and has different sensitivity profile that peaks at around 500nm  photopic= referring to light intensities that are bright enough to stimulate the cone receptors and bright enough to “saturate” the rod receptors (drive them to their maximum response)  scotopic= referring to light intensities that are bright enough to stimulate the rod receptors but too dim to stimulate the cone receptors  STEP TWO: COLOR DISCRIMINATION  Can see detect wavelengths between 400- 700nm, but how can we tell them apart? o Ex. For a single type of photoreceptor which is maximally responsive at 550nm- at the same intensity of light- an equal amount of 450nm of light will produce the same response that 625 nm of light does (BUT 625nm of light looks orange, while 450nm of light looks violet) o Also, wavelength vs. receptor response graph represent a photoreceptors response rate when all wavelengths are presented at the same INTENSITY but decreases in the intensity of the wavelength that the photoreceptor is maximally responsive too (550nm in this case), can cause the same response rate as 450nm of light at a higher intensity o = infinite set of different wavelength-intensity combinations can elicit exactly the same response output of a single photoreceptor cannot by itself tell us anything about the wavelengths stimulating it  problem of univariance= the fact that an infinite set of different wavelength-intensity combinations can elicit exactly the same response from a single type of photoreceptor. One photoreceptor type cannot make color discriminations based on wavelength. o Explains= lack of color in dimly lit scenes (only one type of rod photoreceptor same photopigment molecule= rhodopsin same sensitivity to wavelength can tell light from dark under scotopic conditions, but impossible to discriminate colors nighttime colorblindness= HINTS that color is psychophysical not physical same mix of wavelengths that produces color perception during the day remains present on a moonlit night, but we fail to see colors under dim illuminants because only rods are stimulated= not permitting color vision)  Trichromatic solution o We can detect differences btwn wavelengths/mixtures of wavelengths precisely because we have more than one kind of cone photoreceptor any wavelength from about 420 to 660nm (we can see color from 400-420, and 660-700, but they very long/short wavelengths stimulate only one type of cone) produces a unique set of 3 responses from the three cone types combined signal= a triplet of numbers for each “pixel” in the visual field= basis of color vision o The combined signal (triplet number) also prevents variation due to intensity, because for example a specific light produces a specific set of responses from the three cone types suppose that the light produces 2X as much M response as S response and twice as much S response as L response when the intensity changes, the response sizes change but the relationship doesn’t the relationship between the responses from the different cones defines our response to light and the color we see o Trichromatic theory of color vision (trichromacy)= Young- Helmholtz theory= the theory that the color of any light is defined in our visual system by the relationships of three numbers- the outputs of three receptor types now known to be the three cones  basis of our ability to discriminate one light from another  Almost every light and every surface that we see is emitting or reflecting a wide range of wavelengths:  Metamers= different mixtures of wavelengths that look identical. More generally, any pair of stimuli that are perceived as identical in spite of physical differences.  KEY POINTS 1. nervous system knows only what cones tell it if a mixture of red and green lights produces the same cone output as the single wavelength of yellow light, then the mixture and single wavelengths must look identical 2. mixing wavelengths does not change the physical wavelengths Mix of 500nm and 600nm of light= stimulus is still 500 and 600nm of light NOT average (550nm) and NOT sum (1100nm) color mixture is a mental event, not a change in the physics of light 3. For a mixture of “red” and “green” light to look perfectly yellow, we would have to have just the right red and just the right green. Other mixes might look a bit reddish or a bit greenish. SUMMARY: These examples generalize to any mixture of lights all the lights reaching the retina from one patch in the visual field will be converted into three numbers by the three cone types if those numbers are sufficiently different from the numbers in another patch, you will be able to discriminate those patches if not= those patches will be metamers and will look identical even if the wavelengths are physically different!  History of Trichromatic Theory  Basic theory= established by psychophysical experimentation and theorizing= started with Isaac Newton’s great discovery: o Prism would break up sunlight into the spectrum of hues o Second prism would put the spectrum back together into white o “Rays to speak properly are not colored. In the them is nothing else than a certain power and disposition to stir up a sensations of this or that color” Newton knew that color was a mental event  Maxwell developed a color matching technique that was central to Helmholtz’s work on the trichromatic theory o observer would try to use different amounts of “primary” colored lights (red, blue, green) to match another reference color (Ex. Cyan) o central observation: only 3 mixing lights needed to match any reference light 2 were too few, and 4 were more than needed led to Young-Helmholtz conclusion that 3 different mechanisms must limit the human experience of color  additive color mixture a mixture of lights. If light A and light B are both reflected from a surface to the eye, in the perception of color the effects of those two lights add together. o Additive mixture with paints= Pointillsm (Suerat) = up close you see individually colored dots that may not naturally appear on a face= from far looks like normal skin color  subtractive color mixture a mixture of pigments. If pigments A and B mix, some of the light shining on the surface will be subtracted/absorbed by A and some by B. Only the remainder contributed to the perception. o Ex. Colored filters. Take “white” light that contains a broad mixture of wavelengths pass it through a filter that absorbs shorter wavelengths (“yellow” filter)= result will look yellowish pass it through a bluish filter that absorbs all but a middle range of wavelengths wavelengths that make it through both filters will be a mix that looks greenish (all other wavelengths have been subtracted out of the original “white” light) o An additive mixture of blue and yellow lights looks white because it includes a mix of wavelengths that stimulate the three cones roughly equally  Retina to Brain- repackaging the information:  Cones in retina= neural substrate for detection of lights  Neural basis for discriminating between lights with different wavelength compositions? o Look at differences in activities of the 3 cone types begins in the retina computing differences between cone responses  Nervous system computes 2 differences:  1. ( L-M )  2. ( [L+M] – S )  difference between L and M = considerable info about color particularly well suited to appreciating the differences between amounts of blood in skin blushing and turning pale are useful signals to observe and our specific photopigments may have evolved to help us see those signals  don’t use (L-S) and (L-M) comparisons because L and M are so similar that a single comparison between S and (L+M) can capture almost the same info as both of the above  combining L and M signals= good measure of the intensity of the light (S cones make a rather small contribution to perception of brightness) so theoretically might be wise to convert the three cone signals into three new signals (L – M), ([L + M] – S) AND (L + M)  visual system does something close to this  ** nervous system doesn’t send separate L, M and S signals to the brain= sounds useful but since L and M have such similar sensitivities they are generally in very close agreement (if L says, lots of light coming from location X, M would generally just agree) so doesn’t provide lots of useful data, which is why computing differences/relationships is more useful  Cone-Opponent Cells in the Retina and LGN o Earliest work on combination of cone signals was done on fish o De Valois begun to show that theses sort of signals actually exist in the lateral geniculate nucleus/LGN (structure in the thalamus, part of the midbrain, that receives input from retinal ganglion cells and has input and output connections to the visual cortex) in monkeys  Many ganglion cells in the retina and LGN of the thalamus are maximally stimulated by spots of lights cells with receptive fields with characteristic of center-surround organization (ex. Spot of light in the middle= on, light on the periphery= off)  Similar relationship characterizes color some of these retinal and LGN and ganglion cells are excited by the L-cone onset in their center and inhibited by M-cone onsets in their surround= (L-M) cells= one type of cone- opponent cell= different sources of chromatic info are pitted against each other there are also (M-L), ( [L+M] – S ), and ( S – [L+M] ) cells  support the repackaging of cone signals described earlier  cells excited by light onset can be thought of as (M + L) cells form 3 signals that illustrate the theoretical grounds described  Cone-opponent cell= a cell type- found in the retina, LGN and visual cortex- that in effect, subtracts one type of cone input from another  STEP THREE: COLOR APPEARANCE  Summary of Detection and Discrimination: 3 cones detect wavelength rods make small, important contribution to color vision but only in dim light retina and LGN contain cells that have repackaged information into cone- opponent difference signals that constrain our ability to see differences between regions  So light reaching any part of the retina will be translated into 3 responses, one for each local population of cones after this translation, the rest of the nervous system cannot get any more info about the physical wavelengths of light if light reflecting off two surfaces produces the same set of cone response= the 2 surfaces must and will appear to be exactly the same color= metamers, even if their physical characteristics are quite different  Three Numbers but many colors: o Estimated that with just 3 numbers, we can discriminate the surfaces of more than 2 million different colors many are light variations of what we would consider the same color but even if lightness is ignored we can still distinguish about 26,000 colors  Color terminology o Describe each color in the spectrum using a single number (Ex. wavelength) beyond the spectrum= 3D Color space= the three dimensional space established because color perception is based on the outputs of three cone types, that describes the set of all colors starts with the 3 numbers that come from the outputs of the three cone types (analogous to H, W, L of a cube) EX. (RGB OR HSB) o Achromatic color= referring to any color that lacks a chromatic (hue) component= black, white or grey R equal to G and B: black=0, white= 255, and intermediate for any numbers in between if only R and G are set to 255= yellow o Set of three numbers defining the color in terms of HUE, SATURATION, BRIGHTNES (HSB)  Hue= chromatic (colorful) aspect of color (red, blue, green etc.)= each point on the spectrum defines a different hue  Saturation= chromatic strength of a hue= corresponds to a the amount of hue present in a light= Ex. white has zero saturation, pink is more saturated, and red is fully saturated (bloodred= a saturated red)  Brightness= perceptual consequence of the physical intensity of light= physically intense light of the sun looks brighter than the less intense moon o “nonspectral hues”= not present in the spectrum= can result only from mixtures of wavelengths, not from single wavelengths  Opponent colors  Hering- some combinations of colors are illegal= reddish green and bluish yellow don’t exist.  Opponent color theory (Hering) Young & Helmholtz described trichromatic theory with 3 colors, but Hering’s theory has 4 basic colors in 2 opponent pairs: red vs. green and blue vs. yellow o a black vs. white component formed a third opponent pair but this is a bit different because although you can’t have a reddish green, gray can be described as a blackish white o “hue cancellation” starting light= yellowish green cancel yellow by adding opponent color blue measure amount of blue light needed to just remove all traces of yellow  can do the same for the red-green opponency add just the right amount of red to make the patch of color lose all greenness  organish red + green= yellow, BUT pure red (includes a bit of blue) +pure green= white= hue cancellation  unique hues= any of four colors that can be described with only a single color term (red, yellow, blue, green). Other colors (eg. Purple/orange) can be described as compounds (Reddish blue, reddish yellow). Unique hues, ex. Unique blue, can only be cancelled by its opponent color (no red or green to cancel).  4 unique hues but only three have unique loci on the spectrum  very long wavelengths= red (maybe with a touch of residual yellow)  two crossings of red-green function provide the loci of unique blue and unique yellow  blue-yellow function crosses from positive to negative= unique green ** opponent- process sounds similar but is not the same as (L-M) or ([L+M] – S) need another transformation of the color signals in order to make the cone-opponent signals into color-opponent signals  Color in the Visual Cortex  Cells in the LGN seem to be cone-opponent cells so the transformations that produce color-opponent processes are probably in visual cortex  Single-opponent cells= cone-opponent cells= a cell type- found in the retina, LGN and visual cortex- that in effect, subtracts one type of cone input from another EX. (center= R+, periphery= G-)  Double-opponent cell= color-opponent cells= a cell type first found in the visual cortex, in which one region is excited by one cone type, combination of cones, or color and inhibited by the opponent cones or color (Ex. Centre of cell= R+/G-). Another adjacent region would be inhibited by the first input and excited by the second (Ex. Periphery of cell= R-/G+)  Specialized brain areas for color? o Best evidence= achromatopsia= loss of color vision after brain damage= inability to perceive colors that is caused by damage to the CNS o Color information is transformed in visual cortex but how?  Blobs in V1= not very interested in orientation, but seem interested in color  Blobs send output to “thin stripe” regions of V2 V2 to V4 area that is theorized to respond to color with cells that responded not to wavelength but to perceived color  Adaptation and Afterimages  Adaptation can be color-specific o Afterimages= a visual image seen after the stimulus has been removed if you look at one color for a few seconds, a subsequently viewed achromatic region will appear to take on a color opposite to the original color  First colored stimulus= adapting stimulus= a stimulus whose removal produces a change in visual perception/sensitivity  Illusory color seen afterward= negative afterimage= an afterimage whose polarity is the opposite of the original stimulus. Light stimuli produce dark negative afterimages. Colors are complementary red produces green, and yellow produces blue.  Neutral point= the point at which an opponent color mechanism is generating no signal. If red-green and blue- yellow mechanisms are at their neutral points, a stimulus will appear achromatic. (The black- white process has no neutral point).  opponent color mechanisms overshoot the neutral point due to adaptation which is why the opponent/complementary color is seen. **negative afterimages= are not attributed to JUST the cones or JUST one set of cone/ color-opponent processes. Adaptation occurs at multiple sites in the nervous system.  DOES EVERYONE SEE COLORS THE SAME WAY?  Evidence for YES: o To a first approximation, performance on standard measures of color will be the same o However some differences can occur ex. Unique green can vary from at least 495 nm to 530 nm in different people due to many different factors= for example AGE turns the lens of the eye yellow  Evidence for NO: o 8% of male population and 0.5% of female populations have a form of color vision deficiency=color blindness= malfunction in one/more of the genes coding the three cone photopigments found on x chromosome so if defective males will have it because they only have one x, less likely in females because it is a recessive gene S-cone photopigment is coded elsewhere so S-cone deficiencies are rare o Many different kinds of color-blindness. Determining factors: 1. Type of cone affected 2. Type of defect a. Photopigment of cone type is “anomalous” (different from the norm b. Cone-type missing altogether if only 1 missing= color space becomes 2D flatter color experience but will still have color o M- and L- cone defects are the most common= so most color blind individuals have difficulty discriminating lights in the middle- to long- wavelength range o Deuteranope= an individual who suffers from color blindness that is due to the absence of M-cones EX. 560 and 610nm lights will be classified as the same color o Protanope= an individual who suffers from color blindness that is due to the absence of L-cones different set of color matches based on the outputs of his two cone types (M & S) o Tritanope= an individual who suffers from color blindness that is due to the absence of S-cones o color-anomalous= typically have 3 cone photopigments, but two of them are so similar that these individuals experience the world in much the same way as individuals with only 2 cone types better term for what is usually called “color-blind” most “color-blind” individuals can still make discriminations based on wavelength, but the discriminations will be anomalous/different from the norm True color blindness= few rare forms: o Cone monochromat= an individual with only one cone type. Truly color- blind. See the world only in shades of grey. o Rod monochromat= an individual with no cones of any type, in addition to being truly color blind, rod monochromats are badly visually impaired in bright light. Also have very poor acuity because rods are generally absent in the fovea. **following three forms of color blindness are caused by brain lesions: o achromatopsia= loss of color vision after brain damage= inability to perceive colors that is caused by damage to the CNS caused by lesions in the visual cortex BEYOND the primary visual cortex sees the world as drained of color even though wavelength info is processed at earlier stages in the visual pathway o color agnosia= patient can see something but fails to know what it is o color anomia= inability to name objects in spite of ability to see and recognize them (as shown by usage) inability to name colors in this case  Evidence for MAYBE:  People would agree about the redness of a good example of red, but might be some disagreement about marginal colors reddish orange or orangish red? o Do most languages honor the same color categories? o Number of basic color terms differs dramatically across cultures what makes a color term basic?  Common (like red not BEIGE)  Not an object or substance name (so excludes bronze/lavender)  Not a compound word (blue-grey)  Sounds subjective but Berlin and Kay argue that in English these rules yield a list of 11 terms  Other languages have different numbers of basic terms= can be as few as 2 or 3 o Cultural relativism= in sensation and perception, the idea that basic perceptual experiences (ex. Color perception) may be determined in part by the cultural environment each culture was free to create its own linguistic map of color space  Berlin and Kay’s important discovery= different cultures maps are actually rather similar  11 basic color terms is close to the upper limit words themselves may differ (red, rouge, adom) but languages don’t randomly select possible color names  if a language has only 2 basic color term divided into light/dark  3 color terms one chromatic term, beyond light and dark usually red next that enters (when 4 terms are present) is yellow then green then blue o Experiment: match blue poker chip to another easier to match the next color if it has the same label as the original color, but harder if its different easier to remember which of two colors you had previously seen if the two choices are categorically different even if the distance in color space were the same found that performance of the task reflected the same color boundaries even though the language (which only had 2 basic color terms= 1.light-warm colors, 2.dark-cool colors) didn’t recognize the distinction between two colors leads to conclusion that color perception is not especially influenced by culture and language= blue and green are seen as categorically different even if one’s language does not employ color terms to express this difference o BUT new experiements show that this is not necessarily true  Roberson studies Berinmo tribe from New Guinea limited set of basic color terms BUT terms form novel boundaries in color space (Ex. Distinction in middle of colors we categorize as green which may distinguish live from dead/dying foliage) performed better across this boundary than the blue/green boundary in contrast with English speakers  Also new evidence= learning new color subcategories produces increases in gray matter in parts of the brain implicated in color vision maybe culture does influence color perception  COLOR OF LIGHTS WORLD OF COLOR  so far= detection, discrimination and appearance of isolated lights apply to natural world  colors in the scene can alter the color of a target region:  color contrast= a color perception effect in which the color of one region induces the opponent color in a neighboring region  color assimilation= a color perception effect in which two colors bleed into each other, each taking on some of the chromatic quality of the other  colors in the scenes can also contain colors that cannot exist in isolation  unrelated color= a color that can be experienced in isolation  related colors= a color, such as brown or gray that is seen only in relation to other colors. For example a “gray” patch in complete darkness appears white.  Few thousand unrelated colors context effects allows/boosts number of distinguishable colors to the millions  Color Constancy  Color constancy= tendency of a surface to appear the same color under a fairly wide range of illuminants colors of objects appear relatively unchanged in spite of substantial changes in the lighting conditions  difficult for the visual system to solve heart of the problem= illuminant (the light that illuminates a surface) frequently changes  spectral reflectance function= the percentage of a particular wavelength that is reflected from a surface  spectral power distribution= the physical energy in a light as a function of wavelength relative amount of light at different visible wavelengths ex. Sunlight and skylight  light reflected into our eyes= product of the surface and the illumination 2 different products of surface and illumination are converted into 2 different sets of three numbers by L,M and S cones the two sets of 3 numbers will be different in 2 different lighting conditions but HOW are the colors perceived to be the same?  color constancy is good because we don’t really care about the spectral composition of the light  Color perception= I (illuminant) * S (surface)= given the product of this, the visual system is pretty good at figuring out what S is and relatively what I is too because you can tell the difference between the same color under brighter or darker lighting  Physical Constraints make Constancy Possible o Assumptions made about the illuminant  Natural light sources (and most artificial ones such as standard lightbulbs) are generally broadband= contain many wavelengths even though some are not as intense as other AND their spectral composition curves are usually smooth  This generalization however is violated by some artificial light sources monochromatic light sources make the world look highly unnatural (Ex. In clubs) monochromatic sources make color vision impossible (but very little broadband light is needed to get color back) o Assumptions can be made about surfaces  Real surfaces also tend to be broadband in their reflectance= the percentage of light hitting a surface that is reflected and not absorbed. Typically given as a function of wavelength.  Limits on reflectance= whitest surface rarely reflects more than 95% of any wavelength and blackest surface rarely reflects less than 5%  brightest thing in a visual field is likely to be white  specular reflection= wavelength composition very similar to that of the illuminant (shiny spot on billiards ball) o Assumptions can be made about the Structure of the world  Sharp borders are almost always the result of boundaries between surfaces not between light sources   Shadow borders are an exception: shadow can produce a sharp edge that is unrelated to any change in the underlying surface however the change across a shadow border is typically a change in brightness and not a change in chromatic properties  Visual system uses these assumptions to achieve color constancy  Vision isn’t just a simple translation of world from photons to neurons= IS the nervous systems best guess about what is happening  Experiment by Bloj, Kersten and Hurlbert  Step 1: start with a card, half red, half white.  Step 2: fold it so that red faces white.  Step 3: light reflects from red onto white (making it seem pink)  Step 4: visual system “knows” about reflection and knows how to discount it, so knows that second half is white (perceived as white)  Step 5: fool vision system into thinking the card is folded like a roof  Step 6: without reflection explanation, the white side now looks quite pink  Color Vision in Animals  2 realms of behavior where color vision is especially useful in the animal kingdom= eating and sex= makes it easier to forage for food, or help in mating pick ripe vs. raw, flowers have variations that are designed to attract bees for pollination, colorful displays can be sexual signals to attract mates (as can indicate better genes), primate color vision may be particularly well suited to detect amount of blood in a blushing/blanched cheek  Color vision is accomplished in different ways in different species  Humans= trichromats= 3 different types of photoreceptors, Dogs= dichromats= 2 types of photoreceptors, Chickens= tetrachromats= 4 types of photoreceptors but more than 3 or 4 doesn’t increase info much so generally not more than this  Human S, M , L cones are different because they contain different photopigments BUT it is possible to use a single photopigment to create more than one functional type of cone o Put a different filter in front of each type of cone so that some wavelengths are subtracted before light reaches the photoreceptor Ex. A cone with a reddish oil droplet in front of it will respond more to long-wavelengths of light than cone with greenish droplet (chicks+ birds + reptiles have these droplets)  Different species have different lights and each species visual system appears to be tuned to its particular wavelength signature (fireflies make their own light= bioluminescence)  Humans and other animals= interested in the properties of surfaces, not really the properties of illuminating lights color constancy= goal of all animals with di/tri/tetrachromatic vision research shows that animals with quite different color visions can also show color constancy o Honeybee trained honeybees land on one region of color in order to get food and ignore two other colors do this despite differences in illuminating light  Studies in animals remind us about 2 things: o Color= mental experience not a physical property oil droplets of fireflies, tetrachromacy of chickens etc. o Experiments on color constancy illustrate that even though different organisms have different color vision apparatus, all must solve the same basic problems physics of world don’t change for different species humans and bees both have to ignore the color of the illuminant Chapter 6 Space Perception and Binocular Vision  One fundamental goal of the visual system= ability to perceive and interact with the structure of space  realism= philosophical position arguing that there Is a real world to sense= assumption that the external world exists  positivism= a philosophical position arguing that all we really have to go on is the evidence of the sense, so the world might be nothing more than an elaborate hallucination  Euclidean= referring to the geometry of the world. In Euclidean geometry, parallel lines remain parallel as they are extended in space, objects maintain the same size and shape as they move around in space, the internal angles of a triangle always add to 180 degrees and so forth. o Real world is Euclidean, but the geometry of retinal images of that world is decidedly non-Euclidean the 3D world is projected onto the curved 2D surface of the retinaEx. Parallel lines don’t necessarily remain parallel in the retina, the retinal area occupied by an object gets smaller as the object moves farther away from the eyeball if we want to appreciate the Euclidean world, we have to reconstruct it from non-Euclidean input we generally reconstruct the world from 2 non-Euclidean inputs= the two distinct retinal images o The two retinal images differ because the retinas are in slightly different places but our visual system goes to elaborate lengths to both exploit and reconcile these differences  Advantages of 2 eyes o Evolutionary same as having 2 kidneys, lungs etc. If you lose one eye you can still see. o Enable you to see more of the world rabbits visual field= 360 deg. (make it hard to catch them) humans visual field= 190 def. from left to right, 110 deg. is covered by both eyes more restricted visual field vertically= 60 deg. up from center of gaze, and 80 deg. down o Binocular= with two eyes overlapping binocular visual fields give predatory animals a better chance to spot small, fast- moving objects in front of them (ex. Prey) o Binocular summation= combination (or summation) of signals from each eye in ways that make performance on many tasks better with both eyes than with either one eye alone may have provided evolutionary pressure that first moves eyes to the front of birds/mammals faces o Binocular disparity= the differences between two retinal images of the same scene. Disparity is the basis for stereopsis, a vivid perception of the 3D world that is not available with monocular vision. o Monocular= with one eye o Stereopsis= the ability to use binocular disparity as a cue to depth (putting cap on pen= harder with one eye than both). Not necessary for depth/space perception but does add a richness to perception of the 3D world  Depth cue= information about the 3D visual space. Can be monocular or binocular.  Geometrically impossible for visual system to create a perfectly faithful reconstruction of Euclidean space, given the non-Euclidean input so we use depth cues to infer aspects of the 3D world from our 2D retinal images using basis of retinal images, physics and geometry, each cue provides hint about the likely structure of the space in front of us and the disposition of objects in that space  every view of the world provides multiple depth cues, that usually reinforce each other, combining to produce a reliable/convincing representation of the 3D world occasionally cues can be contradictory (Ex. Escher drawing manipulates cues into an implausible story)  Monocular Depth Cues= depth cue that is available even when the world is viewed with one eye alone (not related to stereopsis) 1. Occlusion  Occlusion= a cue to relative depth order in which, for example, one object obstructs the view of part of another object  Earlier was also a cue to the presence of an otherwise invisible edge  As a depth cue= gives information about the relative position of objects  Arguably most reliable cue only wrong in the case of “Accidental viewpoints”  Occlusion= Nonmetrical depth cue= a depth cue that provides information about the depth order (relative depth) but not depth magnitude (square in front of a small green triangle or large green triangle that’s just farther away)  Metrical depth cue= unlike occlusion, a metrical depth cue is one that provides quantitative information about distance in the third dimension 2. Size and Position Cues  Image on the retina formed by an object out in the world gets smaller as the object gets farther away  Visual system knows this fact of projective geometry implicitly ** Projective geometry= for purposes of studying perception of the 3D world, the geometry that describes the transformations that occur when the 3D world is projected onto a 2D surface. For example parallel lines don’t converge in the real world, but they do in the 2D projection of that world describes how that world is projected onto a surface Ex. A shadow is a projection of an object onto a surface implicit understanding of the rules of projective geometry can be said to undergird many of the depth cues described  Relative size= depth cue= comparison of size between items without knowing the absolute size of either one  Texture gradient= a depth cue based on the geometric fact that items of the same size form smaller images when they are farther away. An array of items that change in size smoothly across the image will appear to form a surface tilted in depth. (Ex. Can create perception of a ground plane receding into the distance)  Relative height= as a depth cue, the observation that objects at different distances from the viewer on the ground plane will form images at different heights in the retinal image. Objects farther away will be seen as higher in the image. (For objects on a ground plane, a more distant image will be projected higher in the visual field) another geometric regularity produced by projective geometry **Texture fields that provide an impression of three-dimensionality are really combinations of relative size and relative height cues multiple cues interact to produce a final perception  Familiar size= a depth cue based on knowledge of the typical size (or typical relationship between sizes) of objects like humans or pennies **occlusion= non-metrical cue that provides only depth order. BUT relative size and relative height cues, especially taken together, provide some metrical information  Relative metrical depth cue= a depth cue that could specify for example that object A is twice as far away from object B without providing information about the absolute distance to either A or B o EX. RELATIVE HEIGHT AND RELATIVE SIZE  Absolute metrical depth cue= a depth cue that provides quantifiable information about distance in the third dimension o Ex. Familiar size could be an absolute metrical depth cue technically if you know the actual size of an object and the visual angle of the object’s projection on the retina, the visual system could theoretically calculate the exact distance from object to eye 3. Aerial Perspective o Visual system knows about PROPERTIES OF THE ATMOSPHERE in addition to implicit knowledge of geometry and learned knowledge of familiar size o Haze/aerial perspective= a depth cue based on the implicit understanding that light is scattered by the atmosphere. More light is scattered when we look through more atmosphere. Thus, more distant objects are subject to more scatter and appear FAINTER, BLUER AND LESS DISTINCT. **short wavelengths (blue) are scattered more than medium and long wavelengths Ex. Sky looks blue 4. Linear Perpective o Linear perspective= a depth cue based on the fact that lines that are parallel in the 3D world will appear to converge in a two dimensional image o Core piece of projective geometry=lines that are parallel in 3D world will appear to converge in the 2D image, except when the parallel lines lie in a plane that is parallel to the plane of the 2D image o Vanishing point= the apparent point at which parallel lines receding in depth appear to converge **like texture gradient, linear perspective can be seen as a special cue of relative size and height cues (assuming lines are parallel= assumption that distance between lines remains constant, but the RETINAL distance between the lines is larger at the bottom of the image and smaller at the top, so the lines must be farther away at the top of the image than the bottom. **like relative size and height cues, linear perspective provides relative NOT ABSOLUTE metrical depth information 5. Pictorial Depth Cues and Pictures o As a group, depth cues discussed above (occlusion, relative size, relative height, texture gradient, familiar size, haze/aerial perspective, linear perspective) are known as pictorial depth cues o Pictorial depth cues= a cue to distance or depth used by artists to depict 3D depth in 2D pictures **Anamorphosis/anamorphic projection= use of the rules of linear perspective to create a 2D image so distorted that it looks correct only when viewed from a special angle or with a mirror that counters the distortion illustrates that the ability to cope with distortion is limited rules of linear perspective are pushed to an extreme so that projection of 3 dimensions into 2 creates a picture that is recognizable only from an unusual vantage point (or with a curved mirror) UNLIKE successful recovery of shapes in most images, the visual system cannot use knowledge about surface orientation to compensate for the distortion 6. Motion Cues o Motion parallax= Nonpictorial depth cue= an important depth cue that is based on head movement. The geometric information obtained from an eye in two different positions at two different times (motion parallax) is similar to the information from two eyes in different positions in the head at the same time (stereopsis)  movement of head can restore sense of depth during monocular vision (which usually requires stereopsis) o Ex. As you look out of the window of a moving train, objects closer to you shift position more quickly than do objects that are farther away o Parallax= term refers to the geometric relationship that when you change your viewpoint, objects closer to you shift position more than objects farther away don’t necessarily need to move your whole body, just moving your head can accomplish the same phenomena o Provides relative metrical information about how far away objects are o Can provide a sense of depth in some situations in which other cues are not very effective o Downside= only works if the hear moves (just moving eyes back and forth isn’t enough) 7. Accommodation and Convergence o Accommodation= process by which the eye changes its focus (in which the lens gets fatter as gaze is directed toward nearer objects) mechanism by which the eye focuses itself to see objects at different distances o Convergence= the ability of the two eyes to turn inward, often used in order to place the two images of a feature in the world on corresponding locations in the two retinal images (typically on the fovea of each eye). Reduces the disparity of that feature to zero. Usually occurs to focus on something close to you. o Divergence= the ability of the two eyes to turn outward, often used in order to place the two images of a feature in the world on corresponding locations in the two retinal images (typically on the fovea of each eye. Reduces disparity of that feature to zero (or nearly zero). Usually to focus on something farther away. o The more we have to converge + the more the lens has to bulge in order to focus on the object the closer the object is o When we focus on objects more than about 2-3 meters away, the lens is as thin as it can get and eyes are diverges about as much as possible so neither cue provides much info but for objects closer than this limit, visual system takes advantage of both cues (con/divergence and accommodation) o Convergence is used more than accommodation o These cues are the only ones BESIDES FAMILIAR SIZE that can tell us the exact distance to an object.  Binocular Vision and Stereopsis o Binocular disparity= differences between images falling on our two retinas= crossed disparity and uncrossed disparity are both depth cues o Stereopsis= impression of 3 dimensionality- objects “popping out in depth”- that most humans get when they view real-world objects with both eyes  Route of binocular disparity to stereopsis= visual system exploiting regularities of projective geometry to recover the 3D world from its projections but this time onto a PAIR of 2D surfaces **remember object of our gaze always falls on the fovea so object will be in the centre of both left and right visual field  Corresponding retinal points= a geometric concept stating that points on the retina of each eye where the monocular retinal images of a single object are formed are the same distance from the fovea in each eye. The two foveas are also corresponding points themselves.  Vieth-Muller circle= the location of objects whose images fall on geometrically corresponding points in the two retinas  imaginary circle that runs through the two eyeballs and the object of our gaze any item on this circle should project to corresponding retinal points o includes the horopter= the location of objects whose images lie on corresponding points. The surface of zero disparity. binocular single vision is possible here. o Objects that lie on the horopter= seen as single objects when viewed with both eyes. Objects significantly closer to or father away from the surface of zero disparity from images on decidedly noncorresponding points in the two eyes and we see two of each of those objects o Panum’s fusional area= the region of space, in front of and behind the horopter, within which binocular single vision is still possible. o Diplopia= double vision. If visible in both eyes, stimuli falling outside of Panum’s fusional area will appear diplopic. o The farther away from the horopter that an object’s image lands on the retina the greater the binocular disparity the greater the disparity, the greater the distance in depth of the object from the horopter= depth cue o Direction in depth= given by the SIGN of the disparity o Crossed disparity= the sign of disparity created by objects in front of the plane of fixation (horopter). The term crossed is used because images of objects located in front of the horopter appear to be displaced to the left in the right eye, and to the right in the left eye.  Always means IN FRONT OF the horopter o Uncrossed disparity= the sign of disparity created by objects behind the plane of fixation (horopter). The term uncrossed is used because images of objects located behind the horopter will appear to be displaces to the right in the right eye, and to the left in the left eye.  Always means BEHIND the horopter  Stereoscopes + Stereograms o Wheatstone invented the stereoscope= presented one image to one eye and a different image to the other eye= used to present dichoptic stimuli for stereopsis and binocular rivalry proved that the visual system treats binocular disparity as a depth cue, regardless of whether the disparity is produced by actual or simulated images of a scene o Free fusion= the technique of converging (crossing) or diverging the eyes in order to view a stereogram without a stereoscope = “the poor man’s stereoscope”Step.1= cross eyes and relax enough to see just 3 sets of images= far left is seen on in left, far right= only in right, and the middle set=fusion of one set seen by left eye and other set seen by the right eye can also do this by diverging the eyes (focusing on a point beyond the plane of the page) and relaxing them Step. 2= bring the middle set into focus by decoupling accommodation and convergence will cause what is seen as misalignment in monocular vision, the opposite misalignments become the binocular disparity and your visual system converts that disparity into a perception of depth depending on whether you used convergence/divergence to free fuse= the disparities will be reversed along with perceived depth o Stereoblindness= an inability to make use of binocular disparity as a depth cue. The term is typically used to describe individuals with vision in both eyes. Someone’s who has lost one (or both) eyes is not typically described as stereoblind.  3-5% of the population lacks stereoscopic depth perception. Usually a secondary effect of childhood visual disorders such as strabismus (misalignment of 2 eyes) o Random Dot Stereograms (RDS)= a stereogram made of a large number (often in the thousands) of randomly place dots. Random dot stereograms contain no monocular cues to depth. Stimuli visible stereoscopically in rand dot stereograms are Cyclopean stimuli.  Cyclopean= Referring to stimuli that are defined by binocular disparity. Shapes that are defined by binocular disparity, and provide no monocular cues.  RDS’s demonstrate that disparity is sufficient for stereopsis  Theory: stereopsis might be used to discover objects and surfaces in the world would be useful to reveal camouflaged objects  Stereo Movies, TV and Video Games o 3D movies make use of stereoscopic photography (records images as seen from two slightly different perspective) and 3D glasses (which separate the images to the two eyes so that disparity provides additional cues to depth) LCD shutter glasses (contain liquid crystals)= enable presentation of images to each eye that are synchronized with the images on the movie/tv/comp. screen o without glasses lenticular printing= images are digitally split and interleaved (left,right,left,right) at a fixed spacing and a special array of many tiny lenses placed on the screen ensures that the left eye sees only the 3D viewing without glasses  Using Stereopsis o Applications go beyond entertainment o Military gest increased info out of aerial surveillance if your view of the ground is stereoscopic but don’t get adequate disparity from distant targets (ground looks flat from far) because you would need eyes that are separated by hundreds of feet= use plane and special camera that takes pics from specific vantage points o Medical applications= reading a mammogram stereopsis can disambiguate the depth of certain structures reduce rate of error  Stereoscopic Correspondence o Correspondence problem= in binocular vision, the problem of figuring out which bit of the image in the left eye should be matched with which bit in the right eye. The problem is particularly vexing when the images consist of thousands of similar features (like dots in Random Dot Stereograms) easier with simple images with different characteristics o Problem is simpler if look at blurred vision of RDS leaves only few low spatial frequency blobs that are easier to match o 2 additional heuristics for achieving correspondence:  uniqueness constraint= in stereopsis, the observation that a feature in the world is represented exactly once in each retinal image. This constraint simplifies the correspondence problem.  Means visual system knows that each monocular image feature (ex. Nose/dot) should be paired with exactly one feature in the other image  Continuity constraint= in stereopsis, the observations that, except at the edges of objects, neighboring points in the world lie at similar distances from the viewer. This constraint simplifies the correspondence problem.  Means that disparity should change smoothly at most places in the image  Physiological Basis of Stereopsis o Most fundamental requirement= input from 2 eyes must converge onto the same cell this does not happen until the striate cortex where most neurons are binocular (cells have 2 receptive fields, one in each eye) in striate cortex neurons, these receptive fields are generally very similar in the 2 eyes, sharing nearly identical orientation and spatial-frequency tuning as well as the same preferred speed and direction of motion so these cells are well suited to the task of matching images in the two eyes o Many binocular neurons respond best when the retinal images are on corresponding points in the two retinas providing neural basis for the horopter BUT: o Many other binocular neurons respond best when similar images occupy slightly different positions on the retinas of the two eyes these neurons are tuned to a particular binocular disparity o Stereopsis can be used both metrically and nonmetrically  Nonmetrical stereopsis= tell you that a feature falls in front of or behind the plane of fixation disparity-tuned neurons of this sort can be found in V2 and some higher cortical areas: some respond to disparity near zero (images falling on corresponding retinal pints), others tuned to range of crossed (near) or uncrossed (far) disparities  Used in VENTRAL/”WHAT” PATHWAY= near/far/categorical info  Metrical stereopsis= because stereopsis is a “hyperacuity” with thresholds smaller than the size of a cone  Used in DORSAL/ “WHERE” PATHWAY  Fine-grained metrical stereopsis isn’t produced by receptive fields that are just a little displaced from each other the receptive fields in the two eyes may differ in “phase” (Ex. 90 deg. phase shift from left-eye cell to right eye cell) so a physical stimulus in the plane of fixation, would not stimulate both of these cells at the same time  Phase shift corresponds to a binocular disparity and could be used to derive depth from the two retinal images  COMBINING DEPTH CUES  Multiple sources of depth cues need to be combined  Automatic cue combination process= “unconscious inference”= cues can be misleading/wrong so our visual system just makes the best coherent guess of 3D space  Bayesian approach= a way of formalizing the idea that our perception is a combination of the current stimulus and our knowledge about the conditions of the world- what is and is not likely to occur. The Bayesian approach is stated mathematically as Bayes’ theorem- P(A|O)= P(A) * P (O/A) / P (O)  Bayes’ theorem enables us to calculate the probability (P) that the world is in a particular state (A) given a particular observation (O)  Bayes’ theorem P(A|O)= P(A) * P (O/A) / P (O) o P(A|O)= probability (P) that the world is in a particular state (A) given a particular observation (O) o P(A)= prior probability of state A o P (O/A)= probability of observation O, given state A o P (O)= probability of collecting observation O  Generally infinite set of possible explanations for a visual scene because size and distances can vary forever  Cues such as familiar size, or the presence of an accidental viewpoint, can influence our decision regarding a scene  Ideal observer= a theoretical observer with complete access to the best available information and the ability to combine different sources of information in the optimal manner. It can be useful to compare human performance to that of an ideal observer. o Ideal observer analysis= idea that if we know the quality of each source of information, we can determine the best possible performance of someone using that information  Illusions and the Construction of Space o Perceiving 3D images from a 2D picture o Ponzo illusion= can’t tell which pair of 2 lines is the same length, generally pick pair in the center two things that are the same size in a picture, the farther one would actually need to be much larger in the 3D world to produce the 2D image  some argue that the Ponzo illusion is not a by-product of depth cues but reflects amore general aspect of the visual system’s response to titled lines= no conclusion  Binocular Rivalry and Suppression o Preceding sections: world often projects images on our two retinas that do not overlap/images fall on non-corresponding retinal points visual system is physiologically prepared to del with these discrepancies via disparity-tuned neurons in striate cortex and beyond o but what happens when completely different stimuli are presented to each eye? Visual system chooses to suppress one image more interesting of the two stimuli is likely to be dominant  most important factor is which stimulus is more salient to the early stages of cortical processing  high contrast= more salient than low contrast  bright= more salient than dim  moving objects= more interesting than stationary ones o binocular rivalry= the competition between the two eyes for control of visual perception, which is evident when completely different stimuli are presented to the two eyes  but is never completely won by either eye or either stimulus regions of dominance grow and fade over time  stimuli for rivalry are actually quite common non- corresponding images on the corresponding points of retinal fields  if two eyes are seeing different images= problem= if we see one elephant in one eye and one elephant in the other, don’t we perceive two elephants?  If the elephant is within Panum’s area, we fuse its two images into a single stereoscopic perception. If it is outside Panum’s area, we normally suppress one of the copies. o Why don’t we see the rivalry? In part, because we aren’t looking. Our attention and eyes are typically directed toward the foveated object/on objects falling roughly on corresponding points in each eye also acuity is bad in the periphery of the visual fields that, even when objects are vying for binocular precedence, the rivalry is quite indistinct.  Rivalry= another way of the visual system trying to come up with the most likely version of the world trying to put monkey together from different images in different eyes instead of suppressing one entire field altogether  Development of Binocular Vision and Stereopsis  Most visual functions start of badly (but not as badly as we like to think) and then improve steadily until they reach adult levels BUT development of stereopsis is surprising infants are essentially blind to disparity until about 4 months of age and then stereopsis appears suddenly, like out of the blue.  Stereopsis is not an all-or-nothing phenomenon just as an individuals acuity is a measure of his ability to resolve spatial detail stereoacuity= a measure of the smallest binocular disparity that can generate a sensation of depth once stereopsis develops in infants, stereoacuity increases rapidly (over about just 2 months) to adult levels= very different time course compared to development of simple acuity  Monkeys= develop stereopsis earlier and faster= 1 monkey week to 1 human month  Why the sudden emergence of stereopsis at 4 months? o Although newborns make convergence eye
More Less
Unlock Document

Only pages 1,2,3,4 are available for preview. Some parts have been intentionally blurred.

Unlock Document
You're Reading a Preview

Unlock to view full version

Unlock Document

Log In


OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


OR

By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.


Submit