Contour integration in peripheral vision reduces gradually with eccentricity

Abstract

Hess and Dakin [Nature 390 (1997) 602; Vision Res. 39 (1999) 947] reported that normally-sighted subjects using peripheral vision (beyond 10°) were unable to detect paths of alternating-phase Gabors embedded within randomly positioned Gabors, but could detect same-phase paths. This result led them to propose a “fundamental difference” between central and peripheral visual processing. While we were able to replicate many of their results, our normally-sighted observers could detect alternating-phase paths beyond 10°. We found that path detection decreased monotonically as a function of eccentricity (0°–30°) for both alternating-phase and same-phase stimuli. As with most visual functions the more difficult path detection condition (alternating-phase) declined slightly faster. The results for the normally-sighted observers could not be explained by poor fixation. Three people with substantial central vision loss (i.e. they can only use peripheral vision) could see both same- and alternating-phase stimuli with eccentric viewing of 13°–17°. Therefore central and peripheral vision appear to use similar visual mechanisms to perform the task, there being no fundamental difference.