Centroid evaluation in the vernier alignment of random dot clusters

Abstract

Clusters of random dots have an advantage over the line elements of a typical vernier acuity task in that their internal light distributions may be easily manipulated without affecting the location of their boundaries. The ability of observers to detect the vertical misalignment of two clusters of random dots was measured over a range of cluster sizes and number of constituent dots. Provided the resultant dot density is high enough to allow interpolation to occur, results demonstrate that the task is performed by analysing the centroid (centre of gravity) of the clusters along the direction of their offset.