Computational tools for extracting, representing and analyzing facial features
Saúl Heredia, Miguel Padilla, Alfonso Gastélum Strozzi, Patrice Jean Delmas, Jorge Márquez · AIP conference proceedings · 2016
In this work we present the development of a software for intuitive and interactive extraction of 3D landmarks on triangular meshes, by employing techniques of scientific visualization, interactive picking, and haptic feedback. The software was employed as a tool for analyzing facial features, over a database of 35 face models from Mexican healthy subjects, acquired with a low cost structured light stereovision system. We aimed to obtain the statistical variation of a set of 19 craniofacial points, representing facial features, along the population under study, in order to determine the representative features, and modes of variation, via Principal Components Analysis. As a result, the mean face shape was approximated, by using interactive cage-based deformable registration. As for interactive picking and haptic feedback, we implemented geometrical algorithms based on an octree representation, for achieving accurate real time collision detection.