Biomimetic Vision Sensors - Engineered Biomimicry - page 22

∆θ

(a) (b) (c)

FIGURE 1.14 The three primary insect vision configurations: (a) apposition eye, (b) superposition eye, and (c) neural
superposition eye. Adapted from Ref. 36 .

crystalline cones, light exiting multiple cones
may fall on the same rhabdom. This effectively
enhances the light-gathering capability of this
insect vision configuration, but it reduces the
effective acuity due to the blurring effect of spa-
tial superposition [41] .
In the neural superposition eye, illustrated in
Figure 1.14 c, one rhabdomere in seven adjacent
ommatidia shares an overlapped field of view
with one another. This results in overlapping,
Gaussian sensitivity patterns for the individual
light-sensitive cells. These overlapped fields of
view provide a motion resolution greater than
that implied by the photoreceptor spacing of the
retinal array, a phenomenon known as motion
hyperacuity [43] .

These three configurations are illustrated in
Figure 1.14 . The primary difference between the
configurations is the way light is routed and pro-
cessed. Because of these differences, each com-
pound eye configuration has its own inherent
advantages and disadvantages. Insect species
are equipped with a specific configuration that
is best suited for the role the insect has in nature
and the activities it accomplishes.
In the apposition compound eye, the rhabdom
is in direct contact with the apex of the crystal-
line cone. Each cone and rhabdom are insulated
by light-absorbing pigments such that light
leakage to lateral, adjacent structures is signifi-
cantly reduced [41] . The individual light-gathering
contributions from each rhabdomere are pooled.
The spatial acuity of the apposition compound
eye is primarily determined by the interomma-
tidial angle ( Δ θ ) described by

1.3.3 Visual Processing

(1.11)

θ = D / R ,

Let us now review the vision-processing mecha-
nisms of various biological species, followed by
a more in-depth view of the common housefly
vision system. It is important to study these pro-
cesses because they inspire the development of
sensors and the processing of their respective
outputs. We look at early processes that occur in
the first several cellular levels and at more com-
plex processing. Our coverage of these topics is

where D is the diameter of the facet lens, and R
is the local radius of curvature of the eye [36, 42] .
As shown in Figure 1.14 a, Δ θ describes the
angular displacement between adjacent omma-
tidia, thus the name interommatidial angle .
The optical superposition eye pools light from
adjacent ommatidia as shown in Figure 1.14 b.
Because there is no pigment between adjacent