Growth Simulation of Facial/Head Model from Childhood to Adulthood

This paper presents a method to simulate the growth of the facial/head model of a person from childhood to adulthood. The method works by generating continuous data of standard landmarks using auto-regressive moving average based on anthropometry of the head/face. The standard landmarks are assigned to the particular head and face, and vertices of the growing head/facial model are calculated by projecting the vertices to the one dimensional subspace that contains measurements of the landmarks. The vertices of the predicted model are constrained to keep the features of the head/face distinct by grouping vertices into clusters and maintaining proportional change of each cluster in different age periods. This new technology has applications in security, ID search (for missing children), health care (such as, evaluation of dysmorphic children) and forensic sketch artistry.