Extracting Anthropometric Parameters from a Scanned 3D-Head-Model

Abstract

Head-related transfer functions (HRTFs), which reflect the interaction between sound waves and human anatomical structures, are core of virtual auditory environments (VAEs). To achieve authentic VAEs, individual HRTFs need to be incorporated in signal synthesis. Accurately measuring HRTF-relevant anthropometric parameters is important in the anthropometry-based HRTF customization method which is regarded as a promising approach to obtain individual HRTFs. Traditional measurement methods using rulers and photography suffer from low accuracy. This paper proposes a new measuring method for extracting HRTF-relevant anthropometric parameters from a scanned 3D-head-model. The proposed method is validated using a representative head model, KEMAR (Knowles Electronics Manikin for Acoustic Research). First, the real head model of KEMAR was scanned by a 3D laser scanner, then length-, angle-, and area-related anthropometric parameters were extracted from the scanned 3D-head-model by using two dedicated softwares, Pro Engineering and Solidworks. Compared with the original design criteria of KEMAR, most of relative deviations are less than 1%, suggesting the proposed method is accurate.