Visual patterns in the perception of abstract and social stimuli
This thesis investigated with regard to the perception of abstract and social stimuli: (1) What constitutes a visual pattern? (2) Whether people possess a proclivity towards one particular pattern type. (3) When is patterning imposed or detected by the visual system? The abstract stimuli consisted of checkerboard patterns and the social stimuli consisted of faces or social groups. Initially the term "pattern" was defined as an image that contains redundant information. This was illustrated by a bias when defining patterns by members of the public towards images that contain both repeated and reflective symmetry, or a low number of possible variants and therefore reduced information content, i. e. more redundancy. Similarly reflective symmetry was identified as a key property in defining faces. The effect of symmetry type on early visual processing was investigated further in a series of backward masking experiments on both abstract and facial stimuli (Chapters 6& 7). The results of the masking experiments suggest a bias during early visual processing for patterns that contain symmetry (i. e. repetition or reflection), or share common fate compared with randomly generated patterns or distorted faces. A top-hemifield and LVF bias was observed in the early detection of patterns. Patterns that take advantage of these properties such as the eyes within the face were suggested as having a perceptual advantage. Patterning appears to be imposed at all stages of visual processing. At early stages of visual processing, repetition (and in the face the eyes) was observed as having an early perceptual advantage over reflection (and in the face the mouth). However at later stages of processing repetition appeared to be processed serially and no longer had a perceptual advantage over reflection (ISIs >42ms). Reflection was suggested as having a perceptual advantage post V1 (ISIs >96ms). Patterning continues throughout a visual scene from the local level to the global level, as such both the human face and human social groups can be perceived as patterns. This was illustrated by a series of experiments investigating the effect of patterning on the perception of images presented in the periphery of a scene (Chapter 8).