Purchase this article with an account.
Kinjiro Amano, Sérgio Nascimento; Influence of natural illumination changes on hue statistics in natural scenes. Journal of Vision 2018;18(10):216. doi: 10.1167/18.10.216.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Surface color appearance is susceptible to changes in lighting. Colorimetric properties in art paintings under artificial light sources are markedly influenced by a change in illuminant from e.g. a standard daylight with correlated color temperature (CCT) 6500 K to a fluorescent or LED lamp with lower CCT of 4000 K (Amano et al., ICVS 2017). It is unclear, however, whether the natural lighting changes in the environment have a similar effect on colorimetric statistics, and whether they are relevant to color vision mechanisms. To address these questions, four time-lapse hyperspectral radiance images of natural scenes were analyzed colorimetrically. The illumination on the scenes was natural daylight of CCT ranging from 3000 K to 7500 K. The natural scenes included distant views of houses and mountains, a rock formation, and the terrace of a house. Scenes were acquired in Portugal between noon and sunset in July and October. Distributions of chroma values were analyzed with hue angle in CIELAB space and in a scaled MacLeod-Boynton space (Webster et al., J Opt Soc Am A, 2000) with a reference illuminant D65. The profiles of the distributions varied across the scenes and the natural illuminants. But the higher chroma values clustered at constant hue angles in both color spaces, corresponding to approximately blue-yellow direction. Even when the reference illuminant was replaced with each of the global illuminants derived from a neutral surface in a scene, the hue angles for higher chroma values were less affected, although the chroma values increased a little with CCTs. These results suggest that natural illumination changes may influence colorimetric attributes in different ways. Thus, hue may be more robust than chroma in the scenes tested, and contributions to the blue-yellow mechanism may be more critical than to the red-green.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only