Abstract
How do effects of crowding manifest themselves when viewing elements of natural scenes? We studied the effects of crowding in central and peripheral vision using suprathreshold discrimination experiments. Observers rated the differences between two 5.2-deg patches of natural images that were presented alone or amongst four flankers. In the central condition the targets were located at fixation, and in the peripheral condition the targets were displayed at 16 degs eccentricity in the lower right visual field. In Experiment 1, the flankers were identical to one another - either the same as one of the target images (SAME) or completely different (DIFF) - and were located at 5.2, 6.6 or 8.2 degs (center-to-center) away from the target. In central vision, small crowding effects were found only at very small spacing distances in the SAME condition. In the periphery, crowding effects were evident in both SAME and DIFF conditions, although they were significantly higher in the SAME condition. Spacing distance between target and flankers did not (or barely) had an effect. In Experiment 2, the DIFF distractors were different to the targets and to each other, and were located at 5.2, 8.2 and 11.2 degs away from the targets. In central vision, there were a very small crowding effect for the DIFF condition at the nearest spacing but none for the SAME condition. In the periphery, crowding remains significant for both SAME and DIFF conditions, but the effects for SAME were only marginally larger than those for DIFF at the smallest spacing distance. These results are consistent with previous crowding research demonstrating: (1) weaker crowding in central vision and (2) stronger crowding when target and flankers are similar. We postulate that the increased crowding in the periphery with similar flankers and small distances is primarily caused by an increased likelihood for mismatched feature comparisons.
Research supported by EPSRC/Dstl (GR/S56399/01 & GR/S56405/01).