Perceived texture segregation in chromatic element-arrangement patterns: High intensity interference

An element-arrangement pattern is composed of two types of elements that differ in the ways in which they are arranged in different regions of the pattern. We report experiments on the perceived segregation of chromatic element-arrangement patterns composed of equal-size red and blue squares as the luminances of the surround, the interspaces and the background (surround plus interspaces) are varied. Perceived segregation was markedly reduced by increasing the luminance of the interspaces. Perceived segregation was approximately constant for constant ratios of interspace luminance to square luminance and increased with the contrast ratio of the squares. Unlike achromatic element-arrangement patterns composed of squares differing in lightness [Beck et al (1991). Vision Research, 32, 719-743] perceived segregation did not decrease when the luminance of the interspaces was below that of the squares. Similar results were obtained for red and yellow, red and green, green and yellow, green and blue, and blue and yellow squares. Perceived segregation based on edge alignment was not interfered with by high intensity interspaces. Stereoscopic cues that caused the squares composing the element-arrangement pattern to be seen in front of the interspaces did not greatly improve perceived segregation. One explanation of the results is in terms of inhibitory interactions among achromatic and chromatic cortical cells tuned to spatial frequency and orientation. Alternately, the results may be explained in terms of how the luminance of the interspaces affects the grouping of the squares for encoding surface representations. Neither explanation accounts fully for the data and both mechanisms may be involved.