Ear Deformations In The Biosonar System Of Horseshoe Bats

An important function of hearing systems is to detect, localise, and identify sound sources in the environment. Head Related Transfer Function(HRTF) in-dicates how animal auditory system encodes sound information received。 The shapes of head and ear conclude the encode pattern. Outer ear diffracts the received sound signal and enlarge the signals from certain direction as a filter. Diffraction process and the final beampatterns are formed by the outer ear's geometry characteristic.The pinna geometry are diverse in different classes because of the variational living environment and needs. The interspecific variability seen in the pinna shapes across mammals may reflect the adaptations in evolution. The various functions that the hearing system of an animal has to perform can impose di-verging and sometimes even conflicting requirements on the pinna beampattern that cannot be accommodated by a single ear shape and the beampattern it pro-duces. The adaptation are not only showed on geometry interspecific but also on behavior in single species.Bats occupies20%species numbers in mammals, wide spread in world and developed numerous individuals. Bats occupy a wide variety of sensory ecological niches and hence face a broad spectrum of hearing tasks associated with active and passive sonar. Some of this diversity has likely been facilitated by adaptations in properties of the biosonar system such as the waveform of the emitted pulses or the different pinna shapes and beampattern that they produce. Horseshoe bats can also use ear movements to adapt the sensing task. Previously, ear movements in horseshoe bats and other mammals with moveable pinnae have been interpreted as reorientations of a static beampattern. There are two kinds of ear movements:rigid and non-rigid ear movements, in which the second kind haven't been specified and researched individually.This thesis worked on the active non-rigid pinna deformations seen in the greater horseshoe bat (Rhinolophus ferrumequinum) and shows it can effect a significant and qualitative change in the beampattern and hence provide the sub-strate for a behavioral level of adaptation in the animals'ultrasonic beampatterns. The methods used in research are: ●Setup and experiments:Colored dots serving as artificial landmarks were painted on the pinna using a mixture of visible bright pigments (nail polish) and an X-ray contrast agent (barium sulfate). The animal was subsequently transferred to a padded holder which restrained the body but allowed for free movements of the head. The movements of the painted pinna of each individual tested were recorded with a synchronized stereo pair of high-speed video cameras (GigaView, Southern Vision Systems with Rodenstock Rodagon50mm-lens with Navitar25mm modular focus block) at frame rates of200or300Hz. Biosonar emissions of the bat were monitored using a bat detector (D1000X, Pettersson Elektronik AB).●3D landmarks reconstruction from stereo vision:Stereopsis is used to re-constructed3D coordinates of landmarks form stereo images recorded by high speed video cameras. Stereo calibration is the basic step to compute the system parameters(intrinsic and extrinsic parameters of each camera, rotation and translation matrix between two cameras). The landmarks are matched on photos catched by left and right cameras separatively. The3D coordinated of landmarks in left and right cameras coordinates systems are triangulated from locations on photos.●Computer tomography to reconstruct static ear shape:The μ CT model of the pinna was obtained for a single (upright or close to upright) posi-tion of each analyzed pinna sample. Cross-sections of the samples were constructed from sets of its X-ray shadow images using a cone-beam vol-umetric reconstruction method (Feldkamp algorithm). The reconstructed model was down sampled by a factor of three to an isotropic resolution of107.34μm.●Deformable ear model:The complete deformable ear model are combined by The three-dimensional shape models derived from the μ CT and the point cloud of the landmark position estimates for the stereo photos.●Acoustic simulation of ear models:Finite-element based numerical acous-tic methods were used to obtain the beampatterns of the deformed pinna shapes. The beampatterns of the shapes in near field are computed from meshed ear models through Helmholtz differential equation. Kirchhoff in-tegral formulation are used to obtain beampatterns in far field.The main results are:●Cyclical nonrigid changes in pinna shape were observed:During each cycle, the pinna transitioned from an upright position to a bent position and back. When bending,the tip of the pinna moved down,outwards and backwards relative to the head. The lateral portions of the pinna moved in opposite directions with the frontal portion moving distally and the caudal portion moving proximally. The maximum deformation component of the displace-ment was typically15%more than4mm which corresponds to of the pinna height and is similar in size to the wavelengths in the strongest harmonic (4.35.7mm) of the biosonar pulse.●Systematic changes in the beampatterns were seen as the pinnae bent and resulted in qualitative differences between the upright and bent pinna ge-ometries. For the upright position of the pinna, the sidelobes of the beam-pattern were weak compared to the mainlobe. As the pinna deformed, the sensitivity in the sidelobes increased significantly:At maximum, the sensi-tivities of bent and upright pinnae differed by a factor of about five. Some of the sidelobes did even surpass the respective mainlobes to become the global sensitivity maximum. When the pinnae returned from their bent state to an upright position at the end of the deformation cycle, the beampatterns also reverted to their previous configuration.●In general, the direction, width, and shape of the mainlobes changed little across frequency whereas for majority of the sidelobes these properties de-pended strongly on frequency. The sidelobe locations overlapped much less across frequency than those of the mainlobes. This frequency-dependence manifested itself either in orderly patterns of change, e.g., in lobe direction or in less easily comprehensible changes that affected either lobe direction or shape.Our results demonstrate that individual horseshoe bats have the additional flexibility to dynamically reconfigure their beampatterns in ways that could suit different sensing tasks. Since intricate ear movements are not limited to bats, the substrate for such a behavioral level of beampattern adaptation may also be available to other groups of mammals. Similar mechanisms could also be inte-grated into technical systems to combine the flexibility of dynamic beampatterns with the simplicity of diffraction-based beamforming.