Abstract
When corners are embedded in a luminance gradient, their perceived salience varies parametrically with corner angle (Troncoso et al, Perception, 2005), as does corner-related BOLD activation in all human retinotopic areas (Troncoso et al, ECVP, 2005). However, the parametric relationship between corner angle and corner salience may not be solely restricted to corners within luminance gradients, but hold true for all types of corners in the world. To test this hypothesis, we developed a variant of the flicker-augmented contrast illusion (Anstis and Ho, 1998): we flickered solid (i.e. non-gradient) corners of different angles from black to white (50% luminance over time) against a black or a white background. Sharp corner angles generated a qualitatively stronger illusory effect than shallow corner angles.
To objectively quantify this effect, we conducted a 2AFC experiment, where subjects had to indicate which of two stimuli was brighter: a Standard or a Comparator. The Comparator was a flickering corner with 14 possible angles: ±15°, ±30°, ±45°, ±75°, ±105°, ±135°, ±180°. The Standard was a non-flickering bar made of multiple luminance segments, pseudorandomly scrambled (each segment had one of 11 possible levels of luminance).
Our results showed a parametric relationship between the strength of the flicker-augmented contrast illusion and the sharpness of the corner angle. We propose that this relationship between corner salience and corner angle is a general and fundamental principle of corner perception, with potentially crucial implications for the brain mechanisms underlying early visual processing of shape and brightness.
These experiments were funded through a startup award from the Barrow Neurological Foundation to SM-C.