In Hoffman's book (2000) called
Visual Intelligence: How We Create What We
See, Hoffman talks about how the brain interprets 2-dimensional images
into 3-dimensional objects. 35 rules are presented by Hoffman on how our
brains use our understanding of math to create these optical illusions. To
understand the mathematics of these optical illusions, I will refer
to and use only a few of Hoffman's rules listed below:
- Rule 1. Always interpret a straight line in an image as a straight line in three dimensions.
- Rule 2. If the tips of two lines coincide in an image, then always interpret them as coinciding in three dimensions.
- Rule 3. Always interpret lines collinear in an image as collinear in 3D.
- Rule 11. Construct subjective figures that occlude only if there are convex cusps.
- Rule 14. Rule of concave crease: Divide shapes into parts along concave creases.
- Rule 15. Divide shapes into parts at negative minima, along lines of curvature.
- Rule 16. Minima rule of silhouette: Divide silhouettes into parts at concave cusps and negative minima of curvature.
- Rule 22. Interpret abrupt changes in hue, saturation, and brightness in an image as changes in surfaces.
- Rule 34. Interpret image motions as projections of 3D motions that are rigid and planar.
Literal Illusions
We technically see the world in 2-dimensional space, but with shadows,
size, and perception we make 3-dimensional spaces in our minds. Literal
illusions happen when we see silhouettes of objects that are like other
objects. This means we can follow rule 16 (Hoffman 2000). If we take the
example of the tree branch, each branch has concave cusps between each
twig and are broken into individual parts at each twig connection. Once
our brain automatically breaks the branch into parts it interprets the
silhouette as a similar silhouette, one we are more familiar with, like
a hand. The same principle is applied when looking at the pile of clothes
in a dark room, we see the silhouette broken into separate pieces and
connects it into a human-like form.
Physiological illusions
Rule 34 states that motion follows planar and rigid translations (Hoffman 2000).
This afterimage effect happens because the stimulus of the image continues
after looking away from the image. For example, with the waterfall afterimage
effect, the rocks climb upwards where the stimulus, the water,
was falling downwards. There is a lot of theory behind why this happens,
it has been conceptualized that it is based on a fatigue of neuron groupings
(Eagleman, 2001). Whatever the reason this occurs, the illusion being viewed
is still created with math. Given rule 34, the motion is viewed as rigid and
planar. The afterimage effect stays on the same plane that the stimulus was
given, and the translation of certain points of stimuli maintain the same
distance. The afterimage doesn't get bigger or smaller than the actual image.
Hermann's Grid
can be explained through rule 2 and rule 22. Through rule 22
we abrupt change of color from the black grid and the white lines means
they reside on different surfaces. By rule two the coinciding white lines
are coinciding in 3D or on the same plane.
This leads to an effect called
lateral inhibition, which causes a bright surround to an area appear darker
and, conversely, a dark surround to an area appear lighter
(Schrauf 1995).
Cognitive illusions
Ambiguous illusions
First looking at
The Young Lady and The Old Lady
the young lady and the
old lady optical illusion.
When we consider the rule 15 (Hoffman 2000), we divide the image
into the concave cusps along the curvature. This would be under the
nose/chin and the eye/wart. When broken into these parts the viewer
then starts to see how the image can look like two things, because
each part can represent different body parts.
Necker's Cube
illusion
utilizes rules 1, 2 and 3 (Hoffman 2000). The lines that create the
cube coincide in the 2-dimensional space and therefore are interpreted
to coincide in 3-dimensonal. We know that the coinciding lines create
a vertex to the cube oriented somewhere in space, then due to rule 3
the points of vertex are collinear, staying on the same continuous line,
but it is up to our minds where on that line it is interpreted. Meaning
it could be coming towards us or going away from us This is why the
phenomenon of the cube
jumping
to different perspectives occurs.
Paradox illusions
The
Penrose Stairs
is impossible in nature because a staircase
goes always up or down. So how can you go in circles? Due to
rule 2 and rule 14 (Hoffman 2000), we can see more clearly why
this paradoxical staircase becomes an illusion. Based on rule 2
if lines coincide with one another they coincide in 3-dimensional space,
therefore since the lines that create the stairs coincide,
occur at the same time, with another all the stairs are connected
to each other. The concave crease of the stairs, based on rule 14,
shows each of the stairs bounded up and therefore each step is
stepping
up
in nature. The
Penrose Triangle
follows rule 2 also
since the lines coincide in the image, therefore they must connect
in 3-dimensional space. The illusion can further be explained by
rule 22, that the change in hue means that there is each surface
is interpreted as a different surface. This means that all the
surfaces are connected and different, which then by their orientation
creates an object that is impossible.
Distorting illusions
The first example, the cafe wall can be explained with rule 2 and rule 3
(Hoffman 2000). We are using the opposite of the rule to understand this
illusion. The black tiles' and white tiles' edges don't coincide with one
another then they don't coincide in 3-dimensional space. Making the
parallel gray lines to appear uneven or not parallel. Since the lines
are interpreted as slanted, then by rule 3 they are also slanted in
the 3-dimensional space.
The next example is
Ponzo Illusion explained
by rule 3 (Hoffman 2000). The edges of the horizontal lines on the
railroad are collinear in the image, and they stay collinear in a
3-dimensional space. Therefore, the line that is longer than the
collinear lines would stay long as it is moved up, and the shorter
line would stay short as it is moved back. This means that the two
lines cannot be the same size in 3-dimensional space, but are in fact
the same length.
Fiction illusions
These illusions create an object when it isn't even there.
Kanizsa triangle
and
Subjective Necker's Cube
can both be explained by rule 11 (Hoffman 2000).
These images are subjective to the viewer because they don’t actually appear
within the image. Both the Kanizsa triangle and the Subjective Necker's Cube
contain contains circles that have pointed concave cusps and therefore
contain the triangle and cube between these circles.