Colour Mechanisms Underlying Visual Search with Heterochromatic Distractors

We studied the colour mechanisms involved in a visual search task. The test stimulus consisted of a coloured target randomly positioned among heterogeneous distractors of two colours. Colours of the target and distractors were specified in the equiluminous plane; a pair of distractors was set to lie on a circle around the target and characterised by the radius, central angle, and chromatic direction of a right bisector of the chord between the pair. The stimulus was presented briefly, and observers were asked to report whether a target was present. Target detectability quantified by d' depended on the central angle and the chromatic direction of the bisector. The central angle affected the detectability of the coloured target but not that of the white one. The coloured-target detectability decreased and reached chance level with increasing central angle from 0° to 180°. For a fixed obtuse central angle, maxima of the coloured-target detectability occurred at two bisector directions, one orthogonal to the target direction and the other along the target direction. This suggests that only two orthogonal colour mechanisms were at play and they changed with the colour of the target. These results and previous findings that the target was detected preattentively when it was linearly separable from the distractors in colour space (D'Zmura, 1991 Vision Research 31 951 – 966; Bauer et al, 1996 Vision Research 36 1439 – 1465) may be explained by the same processes, colour selective filters that linearly combine cone signals followed by peak detectors.