PSYC 212 Lecture 9: Space and Binocular vision
18 Feb 2019
School
Department
Course
Professor
SPACE AND BINOCULAR VISION
FROM 2D TO 3D
Problem w/ space/perception that our visual system must resolve is to perceive a 3D world from
a 2D retinal image
Euclidian geometry = 3D:
• There are rules under Euclidian Geometry that 3D objects follow:
o (1) Parallel lines remain parallel as they are extended in space.
▪ In our retinal image, eventually, these parallel lines will converge
o (2) Objects maintain the same size and shape as they move around in space.
▪ On our retinal image, objects shrink with distance
o (3) Internal angles of a triangle always add up to 180 degrees, etc.
▪ On our retinal image, if the triangle isn’t exactly in our focus, the angles
won’t add up to 180 degrees because the image on the retina is a bit
distorted.
2D info from our retina doesn’t follow Euclidian Geometry → Problem our visual system
must solve
Problem visual system needs to solve: How can we work w/ that 2D image to recompose a 3D
perception of the world. We solve this problem using:
• (1) Monocular Hues
o such as occlusion, linear perspective, shadow, etc. – things that give us the
impression of depth at a local level.
• (2) Binocular Vision and Stereopsis
o when we have two eyes, we have two images of a word, for example, and
there’s a small difference between what we see from both eyes. We can use this
disparity to infer depth.
MONOCULAR CUES
NON-METRICAL VS. METRICAL DEPTH CUES
Document Summary
Problem w/ space/perception that our visual system must resolve is to perceive a 3d world from a 2d retinal image. There are rules under euclidian geometry that 3d objects follow: (1) parallel lines remain parallel as they are extended in space. 2d info from our retina doesn"t follow euclidian geometry problem our visual system must solve. Problem visual system needs to solve: how can we work w/ that 2d image to recompose a 3d perception of the world. We solve this problem using: (1) monocular hues such as occlusion, linear perspective, shadow, etc. We can use this disparity to infer depth. Non-metrical vs. metrical depth cues: an important distinction to make between different types of monocular. Non-metrical depth cues: occlusion, occlusion: a cue to relative depth order in which, for example, one object partially obstructs the view of another object. simplest (and most reliable) cue to infer depth.
