Symmetry and template guided completion of damaged skulls

Abstract

We propose a skull completion framework based on symmetry and surface matching. To obtain reliable symmetry detection on incomplete models, we define a local shape signature consisting of the principle curvature (encoding bending of the surface) and the shape diameter function (encoding volume of the model). Symmetric regions on the skull are used to repair the damages. For big missing regions across both sides of the symmetric plane, we use a template-based completion approach, which first maps the template to the damaged skull, then transplants the region on the template to fill holes. Our proposed completion framework will iteratively conduct symmetry-based and template-based repairs; this significantly improves the completion accuracy and robustness. Several damaged skulls have been repaired successfully in our framework. This skull completion framework will greatly benefit subsequent archeological and anthropological processing and analysis.