Eye tracking: Difference between revisions

From Helpful
Jump to navigation Jump to search
mNo edit summary
mNo edit summary
 
(2 intermediate revisions by the same user not shown)
Line 3: Line 3:
The experimental sort of eye tracking usually
The experimental sort of eye tracking usually
* illuminates the inside of the eye with near-infrared
* illuminates the inside of the eye with near-infrared
* IR camera that  
* IR camera that is zoomed in to look at just the eyes




"Can you do this with a webcam?"
At all? Yes.
Accurately? No.
You can download an app for your phone right now. And its output will be... not great.


Why? Multiple reasons




Line 24: Line 16:


Now rotate your head and keep looking at the same spot.  
Now rotate your head and keep looking at the same spot.  
Your pupils are moving from left to right, the place it should detect does not move at all.
Your pupils are moving from left to right within your eye sockets -- yet the place it should detect does not move at all.
 
From ''just'' the image of the pupil in the eye, you would not get a halfway decent direction.
 
We can fix that by fixing your head in place, and while "please rest your chin here" is fine,
"please allow us to clamp your head in place"... makes people not like you much.
 
Another fix to this is to also track the cornea.
 
Corneal-reflection-style eye tracking (there are others) adds a reference.
That reference is the back of the eye. It is also free to move but, being (roughly) spherical,
whatever the backmost part of the eye is (which is the part that will reflect),
the place of most reflection will barely move when the eye looks around,
so it acts as an anchor.
 
This definitely still has limitations, but is also a good step less janky than just pupil.


From ''just'' the image of the pupil in the eye, you would not get a halfway decent direction
unless we clamp some heads in place.




"Can't you add, I dunno, 3D head tracking?"
"Can't you add, I dunno, 3D head tracking?"


Yes, and that does help, but also adds assumptions to the whole, as well as another limited-accuracy thing .
Yes, and that can help, but at the same time also adds assumptions to the whole, as well as another limited-accuracy thing .
 
 
 
 
 
"Can you do this with a webcam?"
 
At all? Yes.
 
Accurately? No.
 
You can download an app for your phone right now. And its output will be... not great.
 
Why? Multiple reasons
 
 
 
'''"Could I do this with an IR-sensitive webcam?"'''


...plus an IR illuminator?


The corneal reflection style eye tracking adds a reference that is also free to move - the back of the eye.
It would work better than just a webcam, but still not great.


This definitely still has limitations, but is





Latest revision as of 15:53, 29 April 2024