Abstract
Finite element human body model (HBM) positioning remains a challenge and automatic methods are essential to enable their effective use in a wide range of applications such as injury analysis in traffic accidents, sports, and forensic reconstructions. In this study, we present a new HBM positioning framework based on a hard-constrained Laplacian mesh deformation as its core, accompanied by both pre- and post-processing to enhance mesh quality, especially in joint areas, which are often a major source of mesh distortion during positioning. Specifically, the proposed pipeline leverages Blender to generate skin and skeleton surface meshes as target postures. The internal free node positions of the HBMs are then computed via Laplacian-based transformations with hard constraints. Notably, we propose the integration of thin-plate spline radial basis functions (RBFs) as an essential component of the framework to predict the positions of additional constraint nodes and to automatically repair distorted elements following Laplacian transformation during the pre and post processing steps. The performance of the framework was demonstrated through three cases using two HBMs (THUMS and PIPER), which involved substantial posture changes, including transitions from the seated to the standing position. Results show that the proposed framework yields smooth deformations while effectively minimizing mesh distortion. In particular, the inclusion of extra constraints effectively mitigates contact penetration and preserves anatomical fidelity, particularly in regions affected by large joint deformations or involving anatomically adjacent but physically unconnected components. In summary, this framework provides a robust and versatile solution for HBM positioning, offering a new alternative to existing approaches such as simulation-based and RBF interpolation-based methods.