Purchase this article with an account.
Alex Kale, Joris Vincent, Steven L. Buck; Natural Scenes are a Possible Source of Luminance-dependent Long-term Chromatic Adaptation. Journal of Vision 2016;16(4):28-29. doi: 10.1167/16.4.25.
Download citation file:
© 2017 Association for Research in Vision and Ophthalmology.
Our recent psychophysical work (ref 1) reveals different red-green balance points when long-wavelength stimuli appear dark in a bright context than when the same physical stimuli appear bright in a dark context. Dark stimuli look greener; bright stimuli look redder. This suggests a difference in L- and M-cone signal weights in processing of dark and bright stimuli.
The red/green shift might arise from adaptation to regularities in the environment. Past work shows that L/M weights shift with long-term adaptation both under experimental conditions (ref 2) and naturally over the lifespan (ref 3). Perhaps long-term adaptation of L/M weights occurs independently for bright and dark stimuli, driven by luminance-dependent differences in chromatic content. Thus, in the natural world, an abundance of L-cone stimulation in dark regions, or at lower luminances, might be compensated by stronger weighting of M-cone signals than in bright portions, or at higher luminances. We examine what luminance-chromaticity interactions exist within natural scenes.
Our analyses show some regularities in the natural world that could support luminance-dependent chromatic adaptation in the direction consistent with the observed psychophysical shift of red-green balance. Analysis of 238 natural scenes from calibrated databases reveals that bright clear sky is a strong and ubiquitous driver of the predicted interaction. On average, pixels in the most luminous quartile were shifted towards smaller L/M ratio compared to those in the least luminous quartile. This luminance-chromaticity interaction was less common and weaker for natural scenes not containing sky.
This PDF is available to Subscribers Only