Computational Geometry in the Human Brain - Discrete and Computational Geometry and Graphs - page 147

Fig. 3. Stronger version of the Cafe Wall illusion (Sugihara, 2012).
3 Ouchi Illusion and Local Motion Detection
The second example we consider is the Ouchi illusion. The picture shown in Fig. 4
is included in Ouchi's book [ 16 ]. This is a still picture, but the central circular
area seems to drift at random, independently from the surrounding area. This
drift illusion can be explained in the following way.
Fig. 4. Ouchi illusion (adopted from [ 16 ]).
First, our retina, which is an array of photo sensors, usually has some slight
random motion. A sensor generally decreases its sensitivity if it detects the same
signal for a relatively long period of time. This is also true of the retina. In order
to avoid this decrease in sensitivity, the retina tries to use different sensors to
detect a given signal. This is why the retina moves, and hence, when we look at
a picture, the image moves on the retina.
Secondly, neurons in the brain, particularly the neurons used in the earlier
stages of processing, cover only a small area of the retina. Hence, they process