Acoustic characterization of human skin using probabilistic methods

Institut

Lehrstuhl für Akustik mobiler Systeme

Typ

Semesterarbeit / Masterarbeit /

Inhalt

experimentell / theoretisch /  

Beschreibung

In acoustic simulation models such as Head-Related Transfer Function (HRTF) computations, human skin is often modeled as a sound-hard surface. However, this assumption fails in the mid to high-frequency ranges where the skin exhibits partial reflectivity. Accurate characterization of the skin's acoustic properties is challenging due to the limitations of direct measurement methods. Specifically, while the specific acoustic impedance can be measured close to the surface using a pu-probe, direct measurement of the acoustic surface impedance remains infeasible. This thesis aims to estimate the normal surface impedance of human skin by reconstructing the acoustic field quantities on the surface using Near-Field Acoustic Holography (NAH). NAH faces challenges, particularly in the presence of noise, as it is necessary to solve an ill-posed inverse problem for the reconstruction. To address these challenges, a probabilistic Bayesian approach is employed, which offers advantages in the presence of noise and enables quantifying the associated uncertainty of the estimation, thereby improving the fidelity of acoustic simulations involving human skin.

Möglicher Beginn

sofort

Kontakt

Jonas Schmid, M.Sc.
Raum: MW1534
Tel.: +49 (89) 289 - 55125
jonas.m.schmidtum.de

Ausschreibung