Abstract
When adjacent elements have similar orientation, a contour integration occurs (Field et al., 1992). Similarly, as the orientations of stimuli become similar the spatial integration of the orientation becomes easier (Dakin, 2001). In the current study, two experiments investigated the effects of contour integration on perceived mean orientations of Gabors. A set of Gabors, some of which induced vertical contours in specific orientations, was presented in an imaginary circle and points of subject equality (PSE) were measured. In Experiment 1, sixty Gabors of 7 stepwise mean orientations (ranging from 4.5° counter-clockwise to 4.5° clockwise with a step of 1.5°) were presented and participants were asked to report whether the mean orientation of the Gabors was counter-clockwise or clockwise. On 50% trials (contour condition), nine out of the 60 Gabors were aligned to induce a vertical contour with the orientation of -6°, 0°, or 6° from the vertical meridian. On the rest (non-contour condition), 9 Gabors were randomly placed and did not induce a vertical contour. We found that the orientation of the contour biased participants' PSEs into the same orientation, shifting perceived mean orientations of the Gabors to the opposite directions. In addition, using the psychometric functions of the non-contour condition, we predicted each participant's responses in the contour condition, suggesting that mean orientation judgments were made separately from the orientations of the 9 Gabors. Experiment 2 was designed to assess the extent to which mean orientation judgments were affected by 9 Gabors aligned but not inducing contours. Nine Gabors were aligned horizontally but these Gabors were vertically oriented. The pattern of PSE shifts was reversed compared to Experiment 1. The results in the two experiments suggest that Gabors inducing an integrated contour do not contribute to mean orientation judgments.
Meeting abstract presented at VSS 2014