The Biosonar of Bats
(continued)
(continued)
From the shape of a bat noseleaf, we can estimate which region in
space the sound emitted by the bat will illuminate. Likewise, we can
use the shape of the outer ear to predict the spatial sensitivity of
the bat for echoes. The shape of both of these regions (for emission
and reception) depends on the frequency of the sound. Hence we can
draw a boundary (an "iso-surface") in three-dimensional space around
these regions for each frequency. Here surfaces colored with red hues
represent high frequencies whereas blue hues represent low
frequencies.
When using active
sonar, bats also have the option to control the distribution of sound
energy in space. This option seems to be used extensively in bats
which emit their biosonar pulses through the nostrils. In animals with
such nasal biosonar emission, the nostrils are surrounded by elaborate
nostrils which function as outer ears, but on the emission side as
opposed to the reception side. In either case, the bat's brain has to
decode the spatial information generated by the acoustic diffraction
at the outer ear through suitable neural signal processing methods.
We study the acoustic diffraction by the noseleaves and outer ears as well as the neural signal processing which extracts the information it generates. In the study of the diffraction effects we pay particular attention to the different noseleaf and ear shapes which can be found in the approximately 1000 bat species which occur world wide. In the study of the neural signal processing we use computer models as well as actual robotic system implementations to test our predictions on real data.
We study the acoustic diffraction by the noseleaves and outer ears as well as the neural signal processing which extracts the information it generates. In the study of the diffraction effects we pay particular attention to the different noseleaf and ear shapes which can be found in the approximately 1000 bat species which occur world wide. In the study of the neural signal processing we use computer models as well as actual robotic system implementations to test our predictions on real data.
cilia_pm@fz-juelich.de