Perception of surface color in complex three -dimensional scenes
The spectrum of light reaching the eye from a surface depends not only on its reflectance properties but also on the chromaticity and spatial distribution of the illumination. Accurate estimation of surface properties, including color and material, is only possible if the visual system takes this dependence into account. A series of three studies examine whether the human visual system discounts the illumination in three-dimensional environments when estimating the color of matte surfaces. Specifically, behavioral experiments were conducted to address the following questions: (1) Do observers systematically discount secondary illumination originating from nearby chromatic objects when estimating surface color? (2) What is the effect of perceived orientation on the perceived color of matte surfaces in rendered three-dimensional scenes illuminated by a blue diffuse light and a yellow punctate light? (3) Which part of the spatial frequency spectrum of the illumination is utilized by the visual system when estimating the color of matte surfaces?
In each study, a parametric model based on the physics of light-surface interactions of matte surfaces (as they would occur in the respective experiment-specific 3D environment) was derived and fit to observers' data. The model allowed for errors in observers' estimates of the chromaticities of the light sources and/or their spatial distribution (equivalent illuminant model).
The results, taken together, indicate that in three-dimensional scenes, the visual system's estimate of surface color is affected by its interpretation of the geometric layout of the scene; the visual system effectively estimates how light flows within a scene as part of normal visual processing; and, the visual system's spatial resolution of the illumination is well matched to the task it is performing.