A comparison between colour and luminance contrast in a spatial linking task

We report experiments which compare the ability of subjects to employ colour vs luminance contrast as a basis for discriminating the degree of collinearity of random element string pairs. The purpose of the study was to determine the extent to which spatial integration mechanisms could utilize colour contrast. In order to probe directly the processes of spatial integration per se, it was necessary to control for any differences in the efficiency with which the visual system utilized colour and luminance contrast to locate the positions of the individual elements in the test stimuli. To do this we first established the "equivalent" luminance contrast of an isochromatic stimulus which produced equal performance to an isoluminant stimulus in a 2 element per string alignment task. This equated the colour defined and luminance defined stimuli for local positional acuity. We then measured performance for both isoluminant and equivalent luminance contrast stimuli for strings consisting of 2, 4, 8 and 16 elements. This tested for any differences in the processes of spatial integration. For both unmasked stimuli and stimuli embedded in luminance noise, there was no consistent trend favouring either luminance or colour contrast as the number of elements in the stimuli was increased. We conclude that the visual system is able to employ colour contrast as efficiently as luminance contrast for collinearity judgements, thus implicating a general role for colour vision in spatial integration tasks.