Abstract
Some familiar objects have a typical color, such as the yellow of a banana. The presence of such objects in a scene is a potential cue to the scene illumination, since the light reflected from them should on average be consistent with their typical surface reflectance. We studied whether adding a familiar object to a scene improves color constancy. We used a memory color matching procedure. Subjects viewed a square reference patch for 10 seconds. This was presented against a variegated background whose mean chromaticity was that of either CIE Illuminant D65 or CIE Illuminant A. At reference patch offset, the background either remained at D65 (D65 presentation) or switched to D65 (Illuminant A presentation). After a retention interval of 15 seconds, a matching patch appeared and the subject adjusted its chromaticity to match the remembered color appearance of the reference. For each presentation illuminant, the reference patch was presented in two ways. In the “Banana” condition, an image of a banana was added adjacent to the reference patch. In the “Absent” condition, the reference patch was presented in isolation. When present, the banana was rendered under the same illuminant as the background. The banana was never present when the matches were set. We measured the shift in memory matches between D65 and Illuminant A reference presentations. For both “Banana” and “Absent” conditions, the data were consistent with partial color constancy. In addition, there was an improvement in constancy for the “Banana” condition compared to the “Absent” condition, consistent with an effect of familiar object color. The memory matches were fairly noisy however, and the improvement in constancy was small. We plan to replicate the measurements and additional control conditions, to more conclusively determine whether a familiar object effect is the correct interpretation of the data.
Supported by NIH RO1 EY10016 and Nikon Corporation.