Lateral interactions within color mechanisms in simultaneous induced contrast.

The perceived color of a region of visual space is a function not only of the spectral composition of the light incident from it, but also depends on the light incident from surrounding regions. The color contrast induced into a region is a result of lateral interactions between neural mechanisms. These interactions were studied by measuring the induced effect of circularly symmetric spatial sine-waves on a circular central test region. The phase of the surrounding sine-waves was changed uniformly in time, inducing a modulation in the appearance of the test. Observers adjusted the amplitude of real sinusoidal modulation in the test in order to null the induced modulation, and the nulling modulation was used as a measure of the induced effect. Spatial additivity was tested by using pairs of sine-waves of distinct spatial frequencies. The results showed that brightness induction can be characterized as a linear spatial process, i.e. the effects of parts of the surround at different distances from the test are summed, after the effect of each part is weighted by a negative exponential as a function of distance from the test. The magnitude of pure chromatic induction, however, is a result of nonlinear spatial interactions. Thus, these results have implications for the connections between visual mechanisms that process brightness and chromatic contrast.

Influence of shape and perimeter length on induced color contrast.

The magnitude of induced color contrast was measured for tests whose areas, perimeter lengths, and shapes were independently varied. Test shapes were smoothly contoured, multiple-lobed figures generated from unitary Fourier shape descriptors. The shapes had from 3 to 40 lobes and were equal in area to a disk of diameter 2 deg, with perimeter lengths of 1.25, 1.75, 2.25, and 2.75 times the circumference of a 2-deg disk. The surround was a 5-deg disk. The surround was modulated sinusoidally along one of the three cardinal directions of color space around an equal-energy white of 50 cd/m2. The observer nulled the modulation induced into the test by adjusting the amplitude of real modulation in the test. The amplitude of nulling modulation was the measure of induction. The main result was that the amount of induction was similar for all tests of equal area irrespective of the shape or the length of perimeter.