A Markerless Method for Personalizing a Digital Human Model from a 3D Body Surface Scan

Abstract

Digital Human Models (DHM) are used for ergonomic design of products. For instance, vehicle ingress/egress motions are simulated for assessing vehicle accessibility. In order to validate simulations, experiments are often needed implying motion capture and motion reconstruction using a DHM. The first step for motion reconstruction is to create a personalized DHM respecting the anthropometric dimensions of the volunteer performing the task. However creating a personalized DHM from external body shape is not straight forward, because the internal skeleton has to be identified from external body shape. Here we propose a four-step method for generating a personalized DHM which matches a 3D scan. The first step is to clean the scan data and to prepare a DHM and a third body surface template. Then, thanks to the use of the third common body template, the correspondence between the DHM and scan surface points is established, making it possible to calculate the transformation parameters by kriging. From estimated position of joint centers, the internal skeleton is scaled and positioned from a known reference posture to the scan position. The third step is then to attach the surface points to their corresponding skeletal segments. The last step is to check and correct the attached skin points around some joints so as to respect the skin to segment structure specific to a DHM. Compared to the method used in the past by manually adjusting a DHM on calibrated photos of several points of views; the proposed method is operator independent and much less time consuming.

Estimation of body segment inertia parameters from 3D body scanner images: a semi-automatic method dedicated to human movement analysis applications

“…It consisted in obtaining a personalized and structured mesh of the body segments, and estimating their inertia parameter by assuming a constant density (see Figure 1). The first part of the method was adapted from Beurier et al (2015). Beforehand a template mesh was created with the help of MakeHuman and Blender softwares.…”