A composite method for human foot structural modeling

Abstract

A novel method including range-sensing scanning with texture and foot anatomical structure morphing basing on OpenSim is proposed. Palpation of important anatomical landmarks on foot surface was conducted by a physical therapist, and a range-sensing device, Microsoft Kinect sensor, was adopted for the 3D textured model acquisition. 3D coordinate data of the landmarks were measured and harnessed in OpenSim for subject-specific skeletal scaling based on a generic foot musculoskeletal model. The muscle attachment point coordinates derived from an anatomy database basing on sampling from East Asia people were used for muscle modelling. Then the 3D textured foot surface was registered with the morphed anatomical structures so that an integrated foot model was generated. The surface landmark locations were then compared with the corresponding internal bony sites and the errors were calculated to evaluate the accuracy and validity of this method. The potential error sources such as soft tissue thickness and scaling error were also mentioned and discussed. This technique is useful to create individual anatomically accurate human digital models for product design and development.