Chapter 10: Perceiving Depth and Size

Question 1

How does the brain determine depth from a 2D retinal image?

Answer 1

• Visual system uses certain cues from retinal images as reliable (but not completely dependable) as well as oculomotor cues

Question 2

What are the two oculomotor cues used for determining depth?

Answer 2

• Oculomotor = feedback from eye muscles
  1. Convergence = to focus on object, both eyes point toward it. Muscles send feedback about how crossed our eyes are
• Closer objects = eyes turn more inward (greater angle)
• Farther objects = nearly parallel lines of sight (smaller angle)
  2. Accommodation = adjusting the shape of the lens so that light from different distances focusses correctly on the retina. Feedback is sent to brain regarding how contracted the ciliary muscle is.

Question 3

What’s the state of the ciliary muscles when the lens is thin versus when it’s thick?

Answer 3

• Ciliary muscle relaxed = thin lens
• Ciliary muscle contracted = thick lens

Question 4

What is the major setback when using convergence and accommodation as oculomotor cues for depth perception?

Answer 4

• They are only useful for distances less than about 1 metre.

Question 5

What are monocular cues?

Answer 5

• Cues based on retinal images, which provide depth info even with only one open
• Both static (pictorial) and dynamic cues (motion)

Question 6

What are the different kinds of position-based cues?

Answer 6

• A form of static cues
• Partial occlusion/interposition (requires unconscious assumptions)
• Relative height = farther objects tend to be nearer the horizon, same with objects that are considered near eye-level

Question 7

What are the different kinds of size-based cues?

Answer 7

• A form of static cues
• Size-distance relation = the farther an object is, the smaller its retinal image becomes. Retinal image sizes are the inverse size of the magnitude of the distance
• Familiar size
• Relative size = can determine the size and distance of an object with the surrounding objects (don’t have to be the same)
• Texture gradients
• Linear perspective/perspective convergence

Question 8

What are the different kinds of lighting-based cues?

Answer 8

• A form of static cues
• Atmospheric perspective = the farther the object, the more air reflected light must pass through, which is often scattered by things like pollen, moisture, dust etc. (farther images are more hazy)
• Shading = light falls on curved surfaces or edges and causes patterns of shading and shadows
• Cast shadows = location of shadows can indicate distance, as well as disambiguate relative height

Question 9

What are the types of dynamic cues?

Answer 9

• Deletion and accretion = the disappearing and reappearing of objects as they move behind and reappear from other objects
• Motion parallax = when we move nearby objects seem to zip past us while far away objects seem to move with us (closer objects move across the retina more)
• Optic flow = Occurs during forward or backward movement where objects near the fixation point move slowly and then quickly expand outwards as they approach us.

Question 10

What is the major binocular cue?

Answer 10

• Binocular disparity = depth based on the differences in relative positions of objects in the two retinal images
• For closer objects, there’s more of a discrepancy
• Stereoscopic depth perception (i.e., using both eyes)

Question 11

Corresponding retinal points vs. Non-corresponding retinal points?

Answer 11

• Corresponding = Points on the retina of each eye would overlap if the eyes were superimposed on one another
• Non-corresponding = Points that do not overlap if the eyes were superimposed onto each other

Question 12

What’s the horopter?

Answer 12

• An imaginary line that is found along the fixation point (like a radius)
• Any objects that fall on this line would be corresponding points (if not, then they’re non-corresponding points

Question 13

What’s the angle of disparity?

Answer 13

• The distance between non-corresponding points
• A greater angle of disparity indicates greater distance from the horopter (can be closer or farther)

Question 14

Crossed vs. Uncrossed disparity?

Answer 14

• Crossed = Occurs for objects nearer than the horopter. Left of the fovea on the left eye, right of the fovea in the right eye.
• Uncrossed = Occurs for objects farther than the horopter. Right of the fovea in the left eye, left of the fovea in the right eye.

Question 15

What’s great about binocular disparity?

Answer 15

• Can provide info about depth up to approximately 200 meters away
• Can detect differences in depth of 4mm at a distance of 5m away

Question 16

What’s the correspondence problem? How is it resolved?

Answer 16

• How do we know which objects in each retina correspond to the objects in the other retina?
• Solution: Binocular depth cells (also called disparity-selective cells). Found in areas such as V1, V2, V4, MT. Tuned to specific angles of disparity

Question 17

What’s the Ponzo illusion?

Answer 17

• The objects in the image are perceived as different sizes, but physically they are the same size.
• Caused by misapplied size constancy scaling. Perceive the ‘farther’ object as being larger because it is the same size (takes up the same amount of space on the retina) as the ‘closer’ one, so we believe it must be larger.

Question 18

What’s the Muller-Lyer illusion?

Answer 18

• The two vertical lines are the same length even though one has forked ends, making it seem longer.
• Caused by a conflicting cues theory. Our perception of the line depends upon 2 cues: The actual length of the line and the length of the overall figure. Both these cues are combined to form a ‘compromise’ perception of length.
• There’s also a misapplied constancy scaling theory but it’s not very accurate.

Question 19

What’s the Ames’ room illusion?

Answer 19

• An individual looks through a tiny hole in a room so it appears that one individual is significantly larger than the other.
• The room is not a rectangle, but is instead much longer at one end and is angled greatly at one end as well.