Contour integration with colour and luminance contrast

In this study, we consider how colour contrast can be used to integrate form and how it interacts with luminance contrast in the task. The performance of form integration was assessed by measuring the detection of a winding ''contour'' of aligned gabor elements embedded in a background of randomly oriented gabors, using both luminance and isoluminant (red/green) contrast, Performance on the task improves with gabor element contrast, and identical performance for colour and luminance contour detection is achieved at high screen contrasts, showing that colour is able to support a complex form integration task. In a second experiment, we investigate whether colour and luminance contrast can be combined in contour integration by measuring the detection of a path with alternating isoluminant colour and luminance elements. We find that contour detection uses both colour and luminance information cooperatively, but performance is much poorer than would be expected from a single common contour integration process which fails to distinguish the two types of contrast. This suggests that there are specific contour integration processes for colour and luminance. In a third experiment, we measure the effects of variations in colour and luminance contrast on contour detection using elements that combine colour and luminance contrast. We find that varying the colour contrast of elements tends to worsen the detection of a luminance contour, as do luminance contrast variations for colour contour detection, These results suggest no special role for colour in integrating contours, and are discussed with regard to their ecological significance double dagger.