Purchase this article with an account.
Rebecca J Sharman, Elena Gheorghiu; Orientation of pattern elements does not influence mirror-symmetry perception. Journal of Vision 2019;19(10):151c. doi: https://doi.org/10.1167/19.10.151c.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Stimuli used to study symmetry perception often lack orientation information. However, orientation noise decreases the salience of symmetrical shapes suggesting that the two signals may interact (Machilsen et al., 2009, JoV). Here we investigate whether (i) orientation and position-symmetry signals interact in the perception of symmetry (Experiment 1), and (ii) symmetry detection mechanisms are sensitive to the orientation of pattern elements (Experiment 2). Stimuli were patterns made of oriented Gabors arranged either symmetrically about the vertical axis or randomly. We varied the amount of position-symmetry by changing the proportion of symmetrical Gabors and measured symmetry detection thresholds using a 2IFC procedure. Experiment 1 had two conditions: (a) ‘segregated’- position-symmetric Gabors had one orientation and the noise Gabors had a different orientation, and (b) ‘non-segregated’- position-symmetric Gabors were of two orientations in equal proportions, as were the random-positioned Gabors. We varied the separation between the two orientations (0°, 30°, 60°, 90°) in different trials. In Experiment 2, the orientation and position symmetry were manipulated independently. We used (a) position-symmetric patterns in which all Gabors had either the same (asymmetric), mirror-symmetric or random orientations. (b) position-symmetric Gabors had symmetric orientations and noise Gabors had random orientations and positions, (c) patterns with no position-symmetry, and varying amounts of orientation-symmetry. We found that symmetry detection thresholds were not significantly different for the segregated and non-segregated conditions, for all stimulus orientation differences (Experiment 1). The thresholds were also comparable across all conditions containing position-symmetry, but were significantly higher for the orientation-symmetry only (no position-symmetry) condition (Experiment 2). We conclude that symmetry detection mechanisms are not sensitive to orientation information. Although orientation is a dominant visual feature in many visual tasks, it does not affect symmetry detection even when the orientation signals are asymmetric, suggesting that symmetry detection mechanisms are solely reliant on position information.
This PDF is available to Subscribers Only