Perceptual judgments of novel contour shapes and hierarchical descriptions of geometrical properties

Abstract

Previous studies of pattern psychophysics have suggested that a form property such as the number of turns and a structural property such as symmetry were useful cues for perceptual judgments of simple forms. However, it is necessary for complete descriptions of more complex forms to use hierarchical indices reflecting global and local characteristics. In this study, we clarified what geometrical properties contributed to complexity and similarity judgments of novel shapes, and examined differences between the two judgments, using Fourier descriptors as a form property, and symmetropy as a structural property. Global and local unevenness were derived from the amplitude of Fourier descriptors, and the hierarchical representation was found in both judgment data. Whereas complexity judgment was based on local unevenness and global symmetry, similarity judgment was, it was suggested, mainly based on global unevenness and symmetry. Moreover, it became clear that geometrical properties important for complexity data were a subset of those for similarity data. These results suggested that more dimensions in geometrical properties were necessary for similarity judgment than complexity judgment.