Tracking Categorical Surface Colour Across Illuminant Changes In Natural Scenes

How well can categorical colour perception be maintained in natural environments with varying illuminants? To address this question, a colour-naming experiment was performed with colour-monitor images of natural scenes simulated under two different daylights of correlated colour temperature 6500 K and 25000 K. Images were obtained from a set of hyperspectral data to enable the accurate control of illuminant and reflectance spectra. Each scene contained a spherical test surface whose digitally manipulated spectral reflectance coincided with that of a sample drawn randomly from approximately 430 Munsell reflectances grouped into eight colour categories, namely, red, green, blue, yellow, pink, purple, brown, and orange. Observers had to name the colour of the test surface in each image, presented for 1 s, by pressing one of nine computer keys corresponding to the eight categorical colour names and a neutral category. Focal colours were estimated from the peaks of the smoothed distributions of observers’ naming responses in the CIE 1976 (u′, v′) chromaticity diagram. The effect of the illuminant change was quantified by a focal colour constancy index, with values 0 and 1 corresponding to no constancy and perfect constancy. Average levels of focal colour constancy were close to those from traditional measures of colour constancy, but with variation across categories and surface lightness. For blue and purple surfaces levels approached 0.9. For many surface colours, colour naming seems to be robust under illuminant changes and may help to anchor non-categorical judgements of arbitrary surface colours in natural scenes. Introduction The reliable recognition of object and surface colour depends on two perceptual phenomena: colour constancy and categorical colour perception. Colour constancy refers to the invariant perception of surface colour despite changes in the spectrum of the illumination [e.g., 1, 2]. Categorical colour perception refers to the tendency to perceive surface colours as members of a limited number of subsets, as illustrated by the eleven basic colour categories, believed to be an inherent and universal property of human colour vision [3-5], and for which there is some neurophysiological support [6, 7]. The interrelationship between these two phenomena has been investigated mainly with abstract coloured patterns [2, 8, 9]. But it is unclear how well categorical colour perception is maintained in more complex natural environments, where surfaces vary markedly in both spatial and chromatic contents. It has been asserted that colour category boundaries are undistorted by changes in illuminations. Dorothea Jameson, for example, observed “I have no difficulty picking up the green notepad on my desk and not the blue one whether I do it in the daytime or at night with the study light on, or whether the blue and green notepads are resting on the bare surface of the desk, or one is on top of the orange journal, or the pile of papers that I have not yet filed”. [10, p. 42] Such everyday experiences seem to be based on direct and absolute inferences about surface colour. But this is sometimes difficult to quantify in laboratory experiments. Categorical colour perception has often been assessed by colour naming, in which an observer assigns a colour name to a test surface using either an unlimited range of terms or a prescribed finite set [2, 8], whereas colour constancy has often been measured by asymmetric colour matching, in which an observer adjusts the colour of a match surface under one illumination to appear the same as the colour of a test surface under different illumination, according to a given criterion [1, 11-13]. The former represents observers’ direct judgements whereas the latter depends on relative judgements [14, 15]. Both have their advantages and disadvantages [14], but colour naming, in particular, may be