Abstract
Human visual system accomplishes color constancy by discounting an illuminant color, which is estimated with luminances and chromaticities of surrounding colors (e.g. Golz & MacLeod, 2002). It was shown that the brighter surrounding colors had greater contribution to estimation of illuminant color than dimmer colors (Uchikawa et al., in press). However, bright colors in a scene do not necessarily carry information about scene illuminant since there might be some light sources in the scene. The question here is whether the visual system can successfully ignore such surrounding colors appearing in the illuminant-mode to accomplish color constancy. We conducted experiments with the surrounding stimuli that were made to be either in the surface-mode or in the illuminant-mode being independent of their luminances and chromaticities. The stimulus was composed of 60 surrounding hexagons of R, G, B bright and dim colors (1) without gap between neighbor hexagons, (2) with black gap, and (3) with three black segments on each hexagon to be perceived as volumetric cubes. Observers performed achromatic setting (paper match) of the central test patch and evaluated color appearance mode of the brightest surrounding color. The results showed the subjective achromatic points differed among the three stimulus types having the same luminance and chromaticity of surrounding colors. This indicates that, in addition to the luminance balance of surrounding colors, the appearance of surrounding colors have influence on color constancy.
Meeting abstract presented at VSS 2012