QUALITATIVE CUES IN THE DISCRIMINATION OF AFFINE-TRANSFORMED MINIMAL PATTERNS

An important factor in judging whether two retinal images arise from t he same object viewed from different positions may be the presence of certain properties or cues that are 'qualitative invariants' with resp ect to the natural transformations, particularly affine transformation s, associated with changes in viewpoint. To test whether observers use certain affine qualitative cues such as concavity, convexity, colline arity, and parallelism of the image elements, a 'same-different' discr imination experiment was carried out with planar patterns that were de fined by four points either connected by straight line segments (line patterns) or marked by dots (dot patterns). The first three points of each pattern were generated randomly; the fourth point fell on their d iagonal bisector. According to the position of that point, the pattern s were concave, triangular (three points being collinear), convex, or parallel sided. In a 'same' trial, an affine transformation was applie d to one of two identical patterns; in a 'different' trial, the affine transformation was applied after the point lying on the diagonal bise ctor was perturbed a short, fixed distance along the bisector, inwards for one pattern and outwards for the other. Observers' ability to dis criminate 'same' from 'different' pairs of patterns depended strongly on the position of the fourth, displaced, point: performance varied ra pidly when the position of the displaced point was such that the patte rns were nearly triangular or nearly parallel sided, consistent with o bservers using the hypothesised qualitative cues. The experimental dat a were fitted with a simple probabilistic model of discrimination perf ormance that used a combination of these qualitative cues and a single quantitative cue.