When a target consisting of a briefly flashed diagonal line segment is combined with a set of auxiliary lines (the context pattern) to yield the perception of a connected, three-dimensional object, it can be identified more accurately than when the context pattern yields a flatter, fragmented perception. This is the object-superiority effect (Weisstein & Harris, 1974). The present experiments utilized six context patterns that yielded significantly different ratings of perceived depth, and three context patterns that yielded significantly different ratings of perceived connectedness. Observers judged which of two target lines was presented; the line was briefly flashed alone or was accompanied by one of the nine context patterns. The context pattern appeared simultaneously or followed the target by a delay of 30 to 270 msec. Results showed that both the variations in perceived depth and the variations in perceived connectedness produced differences in metacontrast functions. Visual response to different spatial frequencies was also tested by blurring the stimuli — thus diminishing visual response to high spatial frequencies. Results showed that blurring the stimuli had dramatic effects on the temporal functions of accuracy versus delay of the context patterns varying in perceived depth (metacontrast functions): the functions for the patterns rated higher in depth became more like those for the flatter patterns. The differences in the metacontrast functions produced by connected versus fragmented patterns were maintained under image blurring. These results were closely related to the way perceived depth and connectedness ratings diminished as the context patterns were blurred. Our findings suggest that the visual channels that respond to high and low spatial frequencies also have different temporal responses to apparent three-dimensionality and connectedness.
Visual detection of line segments: Two exceptions to the object superiority effect