On the detection of feature points of 3D facial image and its application to 3D facial caricature

Abstract

The paper proposes an automated method for extracting 27 feature points from a 3D facial image. It also shows that 3D facial caricaturing can be realized by such a small number of feature points and that these 27 feature points can be extracted automatically. As a result, it was clarified by comparing a 3D caricature of a 3D PICASSO system with that of a famous caricaturist, Mr. Yoshida, that 3D features of the facial image could be utilized to generate a more impressive facial caricature than the 2D version.