Path drawing with raw gaze coordinates

[alvin10079]

Hi

I recorded raw gaze coordinates to draw path on PNG ,
the path contain saccade and fixation(fix=true & fix=false).

I made two situations,first is to read an article from left to right and top to bottom,
another one is to gaze at 9 points for a moment, and five stars for calibration quality,
but there are two problems I am confused:
1. The path is not smoothly.
https://www.dropbox.com/s/5lqjl777ax8tp ... g.png?dl=0
2. When gaze at a point bottom of screen, the gaze point can't focus on small range.
https://www.dropbox.com/s/ubzfcwmnp5c8k ... s.png?dl=0

I found that these problems taken place bottom of screen.
Is there something I need to pay attention or what settings should I take care?
thanks for suggestion.

[Martin]

Looks like there is a lot of noise for the bottom center & right, typically this occurs then the tracking is not robust. Is this behavior consistent across individuals on your setup? My immediate advice would be to try to elevate or lower the device (or seating) and check if the problem remains. Another thing, were glasses worn when this data was collected?

If this problem remains I'll get you in touch with Javier so we can get a debug session going.

[warbyrd]

A similar problem occurs to me - the center bottom calibration point makes problems often - also wearing glasses makes the data "dirtier" and more "false jumps/saccades" occur.

[marzzzel]

I'd recommend applying this filter to the incoming data and you should be good to go:
http://www.lifl.fr/~casiez/1euro/