Crowding acts beyond the locus of binocular suppression

Abstract

PURPOSE: Crowding refers to the detrimental effect on object recognition caused by the presence of nearby objects. Crowding becomes stronger as the target-flanker similarity increases (Kooi, Toet, Tripathy, & Levi, 1994). Prior adaption to an image suppresses its percept in binocular rivalry. We use binocular suppression to dissociate what is presented to the eye (“eye level”) and what is perceived (“percept level’). Here we ask if the similarity effect on crowding happens at the eye or the percept level. METHOD: Three normal-sighted observers performed an orientation-discrimination task with Gabor patches (sigma of Gaussian envelope = 1/3 deg). A red target was presented to one eye at 5 deg in the lower visual field for 100 ms. Four flankers were presented to both eyes with different colours between the two eyes (red vs. green). The perceived flanker colour was manipulated by the colour of adapting patches (1s in duration prior to the presentation of the target). The flankers could have the same or different colour as the target at the eye and percept levels independently. The orientation threshold for 75% correct was estimated through the 2AFC method of constant stimuli with eight target orientations (0.01 to 30 deg tilt from the vertical line). The strength of crowding was measured by threshold elevation - the ratio of threshold in crowded conditions to threshold in isolated condition. RESULTS: At the “percept level, threshold elevation was significantly higher in the same-colour condition than in the different-colour condition (mean difference = 2.33 +/− 1.71). At the “eye level”, no significant difference was found between the same-colour and different-colour conditions (mean difference = 0.31 +/− 0.35). CONCLUSIONS: Crowding was made stronger by the increased colour similarity between the target and flankers at the percept level, but not at the eye level. Our data suggests that crowding happens after binocular suppression.