Closed onebenclark closed 7 years ago
@shbonita Nick and I looked at this and found nothing obviously wrong with our calibration procedure to cause this. There may be something going wrong at a lower level that we can't change. Our proposed solution is to recognize and ignore noise data as it come's in from the tracker. If only one eye is generating noise, it's x and y are set to the other's x and y. This also fixes a bug where if one eye is closed the gaze is plotted incorrectly due to the averaging of the left and right eye positions.
Included with Pull request #60
The EyeX has a severe stuttering problem. We believe that the problem is noise data that plots points outside of the bounds of the screen for one of the eyes. Most likely (0,0). This may be due to dead zones in the calibration.