Abstract
We propose a combined analytical/statistical method to determine an efficient threshold on the dispersion of estimates of the point of gaze (POG) to indicate a user fixation. The experimental data for this study was obtained with an EyeTech TM3 eye gaze tracker (EGT). The experimental protocol to make the user fixate on pre-determined visual targets was implemented using the C language and OpenCV. Subjects first used the system in a training mode, from which an individualized dispersion threshold was obtained. Our approach was verified by applying the individualized threshold to POG data from a second run, in testing mode, with encouraging results.
Author supplied keywords
Cite
CITATION STYLE
Tangnimitchok, S., O-Larnnithipong, N., Barreto, A., Ortega, F. R., & Rishe, N. D. (2016). Finding an efficient threshold for fixation detection in eye gaze tracking. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9732, pp. 93–103). Springer Verlag. https://doi.org/10.1007/978-3-319-39516-6_9
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
