Cues from a Second Illuminant Hamper Color Constancy

Abstract

Abstract: : Purpose: In surface color perception, cues to the illuminant play an important role in color constancy. The role of a second illuminant was investigated to test whether cues from the second illuminant aid color constancy. Methods: A scene with objects, consisting of two regions illuminated by two separate illuminants and separated by an opaque wall, was rendered using RADINACE. The height of the wall was varied to alter how much information about the illuminant from the other region (i.e., the second illuminant) was available. Four scenes were rendered, depending on the height of the wall: «High’ wall (no cue to the second illuminant), «Mid’ wall (few cues to the second illuminant), «Low’ wall (many cues to the second illuminant), and «No’ wall (continuous mixing of the two illuminants). The four scenes were rendered using Illuminant D65 in both regions. Another four scenes were rendered using Illuminant D65 in one region and Illuminant A in the other region. All scenes were presented with a binocular-viewing setup. A test patch, located on one of the objects, was used for asymmetric matching. Five test chromaticities were used in separate conditions. Results: Color constancy was substantially better with the «High’ and «Mid’ wall than with the «Low’ and «No’ wall conditions. Conclusion: Despite the potential usefulness of cues to both illuminants, cues from the second illuminant hamper color constancy.

Keywords: 360 color appearance/constancy • 362 color vision • 347 chromatic mechanisms