问题
In my eye tracking project, the pupil center is jumping a lot and I don't see it as a fixed point.
What should I do?
My idea is comparing the pupil center and pupil in 2 frames with a threshold but it doesn't help the problem. Another point is camera noise.
What should I do to reduce the noises?
I used the starburst algorithm.
Starburst: A hybrid algorithm for video-based eye tracking combining feature-based and model-based approaches.
回答1:
Eye trackers come with 2 types of noise/errors: variable and systematic error. Variable noise is basically the dispersion around the gazed target and the constant drift or the deviation from the gaze target is the systematic noise. For reference, see the following paper:
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0196348
In your case, its the variable error. Variable error arises due to fatigue, involuntary eye ball vibrations, light, and so on. You can remove it by just filtering the gaze data. However, be careful not to smoothen it too much which might lead to the loss of natural fluctuations inherent in the eye ball.
回答2:
Context
Information regarding eye tracking technologies lacks a lot. However, due to the time (not much) that I am working on this topic, I am raising several issues. One was related to an eye tracking library issue, while the other was related to an alternative to a similar problem. As you can see, this topic is not intuitive and trivial to solve. But do not lose your strength.
Answer
Titled as Eye gaze tracking techniques for interactive applications [1], this paper could answer your doubts. The paper presents a review of eye-tracking technologies and focuses on recent advancements that might facilitate its use in general applications. The authors are also presenting a description of the pupil–corneal reflection technique [2, 3] due to its claimed usability advantages, and show that this method is still not quite appropriate for general interactive applications.
Information
Still, it might be important for you to read some general Usability Testing within the use of eye-tracking technologies. One good article is the one titled as Eye-tracking in Usability Testing, where the author describes issues regarding the camera calibration to be the limiting factor for real-life use. On the same hand, the article titled as UXer’s quick guide to eye tracking describes and advice you how to use it.
About
Just who may concern, in our project we developed two repositories. The first repository called as eye-tracker-setup, was a first attempt to extract information regarding the Tobbi Eye Tracker 4C device. However, we did not succeed that much, so that created another repository called eye-tracker-naive.
References
[1] Morimoto, C.H. and Mimica, M.R., 2005. Eye gaze tracking techniques for interactive applications. Computer vision and image understanding, 98(1), pp.4-24.
[2] Jian-nan, C., Peng-yi, Z., Si-yi, Z., Chuang, Z. and Ying, H., 2009, May. Key techniques of eye gaze tracking based on pupil corneal reflection. In 2009 WRI Global Congress on Intelligent Systems (Vol. 2, pp. 133-138). IEEE.
[3] Morimoto, C.H., Koons, D., Amir, A. and Flickner, M., 2000. Pupil detection and tracking using multiple light sources. Image and vision computing, 18(4), pp.331-335.
来源：https://stackoverflow.com/questions/51627016/pupil-center-is-jumping-a-lot-in-real-time-eye-tracking