Abstract
It is well known that "20/20 vision" corresponds to a minimum angle of resolution of 1 arcmin, and some observers have much finer resolution with MARs as small as 0.5 arcmin. For much coarser repetitive patterns, however, observers (who are not permitted to use an external pointer) have great difficulty counting elements that are separated by 4 - 9 times limits of visual acuity. Why is Counting-By-Eye (CBE) so difficult? Landolt (1891) asserted that the difficulty results from limitations in oculomotor control, but later research failed to confirm this plausible theory. For example, based on high-resolution recordings of saccadic eye movements, Kowler and Steinman (1977) concluded that perceptual confusion can be more important than oculomotor factors. Building on such earlier work, we conducted three psychophysical experiments to investigate the effects of spatial structure and reduced luminance on CBE. In all experiments, participants counted the elements of repetitive patterns with and without the benefit of an external pointer. Spatial frequencies of the patterns varied from 1 to 8 cy/deg. Results from one experiment showed that CBE performance is significantly better for 1-D (a barcode) than for 2-D targets (tree rings and concentric circles), although subjective ratings of difficulty were opposite actual differences in performance. A subsequent experiment compared the effects of reduced luminance (100 to 1.0 cd/m2) on CBE, with parallel changes in visual acuity (VA) and contrast sensitivity (CS). Surprisingly, CBE accuracy for both 1-D and 2-D patterns showed no effect of reduced luminance, despite the expected (large) decreases in VA and CS. These findings suggest that (1) the ability to count-by-eye is not limited by mechanisms of the foveal resolution and CS; and (2) with further refinement, the CBE method may provide a fast and simple test of the coordination of visual perception and action.
Meeting abstract presented at VSS 2014