Vision and the Mind

* Imaging Science / Psychology crossover

* Professor Pelz: The role of eye movement - active vision

* "Optical illusions" really "neural illusions"

* Some illusions easy to explain; some aren't

* Very young children see some of these illusions differently; older people have visual experience

* After 20-second video of spiral, you see the spiral's movement pattern even after it's turned off. So, even for a short timeframe, your eye was 'trained' to that movement.

"Rods" and "cones" light0sensitive things in back of retina named for their shapes

"Bipolar cells" feed signal into "ganglion cells", which in turn send the signal to the optic nerve.

Light actually flows through the ganglion and bipolar cells, hits the rods and cones, and then bounces back

Photon absorbed by rod or cone; voltage induced; signal gets changed at the synapse (lateral inhibition: one full-strength positione charge, two half-strength negative charges

Sometimes neurotransmitters are struck by multiple such charges
.

If the input field is 100% uniform, the charges would entirely cancel each other out. However, at edges of objects, the strength of incoming light is different, so only the edges of the visualized object are accentuated.
This is a data-compression scheme; computer image-compression algorithms actually do something similar.

Your visual system

Herman's Grid - black squares, thick white gridlines
Criak-Obrien/Cornsweet illusion - rectangle that appears to have dark gray left half and light gray right half; really a uniform color

Fourier transform: math algorithm that looks for "just the edges"

Blurry image: has no clear edges

Snakes illusion- apparent motion when going from the spiral pattern to a blank surface.

When surface is darker, your optic system integrates the images slower, so you'll see elss flicker on a darker screen

Vertical angle * horizontal angle = degrees-squared on the field of vision

Visual system focuses on detecting motion

Field-of-view vs. acuity compromise

Cones: High acuity [and color vision]
Rods: Low acuity and high sensitivity

If you hold a finger out at arm's length, the fingernail is about 1 degree wide.

There is a blind spot at a certain degree of the field that comes form the space where the optic nerve comes back out

Small blind spot aswell form the following: No rods directly ahead (0 degrees); only cones in that area

little depression at the center of your retina

Limiting acuity in the periphery is a part of the compromise; but need to move the eye around so that the strong-acuity area can cover different parts of the world

Ocular motor system: six muscles in three pairs (one does left-right, one pair does up-down, and one does rotation)

How much info is enough?
How quick is "just in time"?

saccades = rapid movement of eyes to look at a new object or region

Smooth Pursiut: stabilizes view of a moving object

Tremor, drift and microsaccades when you try to hold your eye steady

optokinetic response / optokinetic nystagmus

Have to serialize your visual-information intake
That presents problem of temporal-integration

Various devices to track eye movement in the laboratory
* Measuring reflection off the eye, in some way, is a common method

* Wearable eye tracker is the next development goal; starting to have success on that