Individually tracking both eyes

[rudi]

First of all, great work!
I'm quite new at this, so I apologize if my question is a bit naive or plain dumb.

I'm working on an app that needs to track both eyes independently, for my understanding, this data is found on the FRAME object as the following:

Code: Select all

frame->lefteye->x
frame->lefteye->y
frame->righteye->x
frame->righteye->y


Analyzing this data I noticed sometimes the positions of the eyes differ, sometimes they differ a lot (ie: 100px, around 1 inch or even more).
Shouldn't the eyes, at least on a person with no eye conditions, point at more or less the same position in the screen?
Am I missing something here? Like any other data that may infer on this?

Thank you!

[mathieujofis]

I was working with the individual eye data as well until I realized it is a little buggy (C++). Every once in a while the fixation location would move ~100-200 pixels in some random direction. I changed my implementation to use the aggregate (both eyes) data, and the issue went away. I know this doesn't help your situation much, but it might be worth checking if this is the same issue.