Automatic 3D Face Recognition Using Shapes of Facial Curves

In addition to facial textures, manifested in facial images, shapes of facial surfaces are also important in face recognition. We study facial surfaces by approximating them with indexed sets of level curves of a continuous function, such as the depth function, defined on these surfaces. Level curves of depth function are extracted efficiently using standard techniques on range images. The shapes of facial surfaces are then analyzed implicitly using shapes of these level curves, called the facial curves, using a differential geometric approach that computes geodesic lengths between these closed curves on a shape manifold. The metric for comparing facial surfaces is a composition of the metric involving individual facial curves. These ideas are demonstrated in the context of face recognition using the nearest-neighbor classifier.