Comparison of the Accuracy in Identifying the Geometric Center of Figures with Different Spatial Structures

This study examined the accuracy of locating the geometric center of visual stimuli, represented by different parts of the contour, using eye-tracking equipment. Two methods of stimulus representation were employed. In one, geometric shapes – rectangles, rhombuses, pentagons, and horizontal bars – were composed of lines 2 px thick (1 px = 1.8 arcmin) with a brightness of 300 cd/m ² on a grey background of 25 cd/m ². In the other, the shapes were marked by bright spots 5 px in diameter positioned at the corners. The experiment included three sizes of shapes: large (800 × 400px or 800 × 800 px), medium (400 × 400 px), and small (200 × 400 px or 200 × 200 px). Method Stimuli were presented singly in random order on a 117 cm diagonal, 1920 × 1080 px resolution “Samsung LE 46 B 652” screen. The participant was seated 1 m from the screen, with head movements stabilized using a mount. Eye movement tracking, stimulus presentation, and data collection were performed using the ViewPoint PC-60 system (Arrington Research, Inc.). After calibration, the participant viewed each stimulus for 6 seconds, during which their eye movements were tracked and displayed in real-time. Seven university students, with an average age of 26 years, took part in the study. Analysis Data from the final second of each of the 11 experimental runs were collected to estimate the perceived location of the centroid. Analyses and visualizations were carried out using Python (custom scripts) and MS Excel. According to the average data of all observers, differences in the representation of geometric shapes did not significantly affect the accuracy of center finding. For large rhombuses represented by lines, compared with same figures predicted by corners, the center difference in the X axis was 14 pixels, and along the Y axis – just 3 pixels; for medium shapes, the X difference was 44 pixels, Y – 13 pixels; for small shapes, the X difference was –31 pixels, Y – 4 pixels, for large rectangle X – 19, Y – -21; medium X – -11, Y – -3; small X – -10, Y – 40, for large pentagon X – 35, Y – 19; medium X – -21, Y – -8, small X – -8, Y – 3, for large horizontal bar X – -11, Y – 1, medium X – -82, Y – -10, small X – 43, Y – -3. The variations in perceived locations of the geometric centers of differently depicted shapes were not statistically significant.