Abstract
We can observe clear dark shadows cast by objects outside in a fine day. It has been pointed out that an illuminant in a shadow is likely to be blueish due to the blue sky whereas that in the sun is white due to the direct sunlight. However, we hardly notice that objects appear blueish in a shadow. Color constancy seems to work separately well in a shadow and in the sun. In this study we examined how well illuminant colors were estimated in a shadow. In experiments, we placed two LC projectors in the top-left and the top-right side to make illuminants simulating the sun and a sky, respectively. The left illuminant lit a black-paper cup, set on a background paper with 6 colors, to produce its cast shadow. The right illuminant adjusted the shadow chromaticity. An LCD monitor illuminated two 2.6-deg hexagon holes, made open in the background paper, from behind so that the two holes, served as test stimuli, appeared in the surface-color mode. One hole was made in the shadow and the other in an area directly lit by the left illuminant. Observers adjusted the chromaticity and luminance of the test stimuli so that they appeared as full white papers (paper match) under test illuminants. The results showed that illuminants were estimated in a good degree in the shadow of different chromaticities (blue, yellow, pink, green and white), indicating good color constancy holds in a shadow.
Supported by JSPS KAKENHI Grant Numbers JP18H03247, JP17H01809.