Abstract

What constitutes a perceptual object? A lamp is seen as a lamp only because we have learned the relationship between the base and the shade, but the base might be seen as a perceptual unit without prior knowledge. I suggest that separating perceptual objects from learned objects, or at least creating some continuum along this dimension, can advance our understanding of object perception by identifying the underlying sources of information. As a first step toward this, I propose that the accuracy with which the size of a region can be judged in a brief presentation serves as an objective measure of the perceptual "objectness" of the region. This measure is robust across scale and insensitive to such features as fine surface structure, supra threshold contrast, and contrast polarity, but it depends on the relationship between the elements defining the region of interest and those defining the background. I will relate the results of some previous findings on size judgments to this measure and will report the results of new experiments that further the idea.

© 1992 Optical Society of America