It is well known that a staircase luminance profile is not seen veridically, but appears as the scallopy-like Chevreul illusion. We have shown that adding thin lines (either light or dark) to the centre of each step creates an illusory brightness change at the point of the line. The regions between the added lines and the edges seem to be uniform, with a clear change in brightness at the point where the line was added. The conditions under which the illusion occurred were measured systematically, both by contrast matching and by annulment. One model that can readily account for the illusion is the local-enegy model of feature detection (Morrone & Burr, 1988 Proceedings of the Royal Society of London B, 235, 221-245). Adding the bar to the step creates a peak in local energy at all scales. At the higher scales, the phase of the energy is near zero, the signal for a line; but at the lower scales the phase is near pi/2, the signal for an edge. We propose that the edge signal of the lower scales causes the brightness illusion and that this brightness difference is structured by the feature defined sharply by the higher scales (even though that feature is not an edge). As well as predicting the existence of the illusion, simulations with the energy model predicted quantitatively the apparent contrast of the illusion as a function of stimulus contrast, bar-position and high-pass filter frequency.
ILLUSORY BRIGHTNESS STEP IN THE CHEVREUL ILLUSION
MORRONE, MARIA CONCETTA;
1994-01-01
Abstract
It is well known that a staircase luminance profile is not seen veridically, but appears as the scallopy-like Chevreul illusion. We have shown that adding thin lines (either light or dark) to the centre of each step creates an illusory brightness change at the point of the line. The regions between the added lines and the edges seem to be uniform, with a clear change in brightness at the point where the line was added. The conditions under which the illusion occurred were measured systematically, both by contrast matching and by annulment. One model that can readily account for the illusion is the local-enegy model of feature detection (Morrone & Burr, 1988 Proceedings of the Royal Society of London B, 235, 221-245). Adding the bar to the step creates a peak in local energy at all scales. At the higher scales, the phase of the energy is near zero, the signal for a line; but at the lower scales the phase is near pi/2, the signal for an edge. We propose that the edge signal of the lower scales causes the brightness illusion and that this brightness difference is structured by the feature defined sharply by the higher scales (even though that feature is not an edge). As well as predicting the existence of the illusion, simulations with the energy model predicted quantitatively the apparent contrast of the illusion as a function of stimulus contrast, bar-position and high-pass filter frequency.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
