| Here we simulate a black triangle in front of white lines. In one case,
the triangle is stationary and the lines rotate. In the other case, the
triangle rotates and the lines are stationary. You should see a difference:
the rotating triangle appears as a deforming blob, wheras the stationary
triangle is more readily seen as such (Bruno
& Gerbino,1991). In these, as in many other
phenomena, the visual system uses spatiotemporal information to recover
contours that would otherwise be invisible. However, this process yields
veridical results only if certain motion components are available. In this
case, when the lines rotate there is a local component that translates
along the occluding edge of the triangle. This component seems critical.
In its absence, the visual system is unable to recover the triangle from
the accretion/deletion of the lines. |
|
