3D reconstruction of skeletal mesh models and human foot biomodel generation using semantic parametric-based deformation

ABSTRACT

This paper introduces two methodologies for the reconstruction of skeletal mesh models and for the generation of parametric bone models of the human foot. First, a new Parametric-Based Deformation methodology is presented, which deforms a template 3D mesh according to user-defined semantic parameters in order to derive a new bone model. The proposed method allows the parameterization of dense and detailed 3D meshes of human foot bones. An extensive evaluation is performed to examine the accuracy, quality, and performance of the proposed method with human and animal bones, and versus existing statistical-based methods in the literature. Also, a novel methodology is proposed for the generation of parametric bone models of the human foot using non-calibrated biplanar X-rays. The proposed method has been applied for the production of detail 3D foot biomodels from an initial template biomodel and has been evaluated using human and animal bones.

KEYWORDS:

• 3D bone reconstruction
• parametric-based deformation
• skeletal models
• human foot biomodel
• semantic parameters

Acknowledgments

This research has been co-financed by the European Social Fund and national funds through the Operational Program Education and Lifelong Learning of the National Strategic Reference Framework (NSRF) – Research Funding Program ‘ARISTEIA.’ The authors wish to thank Prof. Ajay Joneja of Hong Kong University, for providing the dataset of animal bones.