EFFECTS OF ORIENTATION ON SYMMETRIC OBJECTS DETECTION ON NOISY BACKGROUND

Image contours correspond to objects' outlines, surface orientation changes, or texture features.Previous research indicates that detecting contours defined by separate elements depends on their closure and curvature.Closed contours are recognized more quickly, but this advantage diminishes when the contour curvature changes.However, more information is needed about how object recognition interacts with symmetry detection and the dynamics of this interaction.The present study aims to investigate the dynamics of detecting equilateral triangles embedded in noise, depending on their orientation.The stimuli consisted of 1,488 Gabor elements arranged in a hexagonal grid with 48 columns and 31 rows.The orientations of the elements were randomized, except for twelve elements aligned along the sides of an equilateral triangle, with a top element perpendicular to the triangle's base.The triangle was positioned randomly in successive trials, and its orientation could be left/right or up/down.Participants were required to identify the triangle's orientation by pressing a joystick button.Thirty-seven healthy observers (mean age 41 years, range 20-69 years; 23 males, 14 females) participated in the study.The temporal threshold for triangle detection was examined separately for the horizontal and vertical orientations.The results of our study revealed significant individual differences among subjects and a shorter detection time for left/right oriented triangles compared to those oriented up/down.The modeling results, obtained through a Bayesian hierarchical model, indicated significant differences in the probability of a correct response based on the orientation of the stimuli.The highest percentage of correct responses occurred when the triangle's apex pointed to the right, while the lowest -when the triangle's apex pointed downward.These results could not be explained by perceptual asymmetries related to the location of objects in the visual field or the orientation of the symmetry axis.They suggest that object recognition occurs in allocentric coordinates and precedes symmetry detection.