Open-source Software for Determining the Dynamic Areas of Interest for Eye Tracking Data Analysis
Leszek Bonikowski, Dawid Gruszczyński, Jacek Matulewski
Abstract
There is a vivid need for an effective, universal and easy-to-use software tool supporting visual attention analysis based on eye-tracking data using the areas of interest for objects moving within the system of coordinates in which gaze position is recorded. Thus, we decided to design such software using existing algorithms offered by the discipline of computer vision, with an intention the eye-tracking research community could freely use it. According to the tests carried out using two datasets, the MOTLD algorithm had the best efficiency among all implemented in the presented software in determining the dynamic areas of interest. Therefore, we recommend it as the default one. This paper presents the first version of the tool and is an invitation to collaborate on developing this software.