Northern Kentucky University
Eye-tracking Data Analysis with Consumer-Grade Webcam
Grade Level at Time of Presentation
Sophomore
Major
Computer Science
Minor
Neuroscience
2nd Grade Level at Time of Presentation
Sophomore
2nd Student Major
Computer Science
2nd Student Minor
Neuroscience
Institution 23-24
Northern Kentucky University
KY House District #
4
KY Senate District #
24
Faculty Advisor/ Mentor
Nicholas Caporusso, PhD
Department
Department of Computer Science
Abstract
For our research project, we worked with a team to develop a software that utilizes consumer-grade webcams to collect eye-tracking data both accurately and accessibly to be applied in various fields, such as concussion diagnosis and EMDR therapy. Specifically, the purpose of our research was to analyze the data collected from the eye-tracking software to extract various data, such as the movement of the eyes jumping from one fixation point to the next or the movement of the eyes following a consistently moving visual stimulus. To reduce errors in the eye-tracking data, we first sought to remove the data containing blinks from the dataset. Using the position of the patient’s eyelids in each frame captured in the dataset, we calculated the three-dimensional area of the eye visible in each frame to determine when a blink occurred. In our future work, we will focus on further processing the data collected to begin extracting other features, such as the velocity of the eyes as they shift from one point of fixation to another to be applied in various uses.
Eye-tracking Data Analysis with Consumer-Grade Webcam
For our research project, we worked with a team to develop a software that utilizes consumer-grade webcams to collect eye-tracking data both accurately and accessibly to be applied in various fields, such as concussion diagnosis and EMDR therapy. Specifically, the purpose of our research was to analyze the data collected from the eye-tracking software to extract various data, such as the movement of the eyes jumping from one fixation point to the next or the movement of the eyes following a consistently moving visual stimulus. To reduce errors in the eye-tracking data, we first sought to remove the data containing blinks from the dataset. Using the position of the patient’s eyelids in each frame captured in the dataset, we calculated the three-dimensional area of the eye visible in each frame to determine when a blink occurred. In our future work, we will focus on further processing the data collected to begin extracting other features, such as the velocity of the eyes as they shift from one point of fixation to another to be applied in various uses.