Method of Characterising 3D Faces Using Gaussian Curvature

This study investigates a new direction of characterising 3D face based on Gaussian curvature for classification problems. This method relies on the geometric facial models regarding as input information. Within a sub-surface we define, which includes much invariant over facial expressions changing, symmetry characteristic is evaluated combining with the study of two distinguishing profiles based on Gaussian curvature. The two profiles have distinguishing features which are classified by means of their curvature. This method is an extension of our previous work on other two profiles feasibility evaluation. A series of curvature results of the profiles are discussed along with the shape of curvatures plot. The experimental results show that the shape information obtained from the 3D facial model can be potentially used to classify and identify individual facial surfaces. In addition, the future work of optimizing and completing this method is presented.