Shape-based averaging for craniofacial anthropometry

We present in this work original methods for defining and obtaining robust, shape-based averages of any feature of the human head anatomy from a normal population. Our goal is the construction of computerized atlases with representative facial features and craniofacial anthropometry for specific human populations. A method for true-morphological averaging in three dimensions is presented, consisting of a suitable blend of shape-related information for N objects to obtain a progressive average. It is made robust by penalizing, in a morphological sense, the contributions of features less similar to the current average. The morphological error and similarity, as well as penalization, are based on the same paradigm as the morphological averaging. We present current results for the shape of the ear in 24 subjects.