3d Face Recognition from Range Image

MIAO, SHUN. 3D Face Recognition From Range Image. (Under the direction of Professor Hamid Krim). In this thesis, we explore the statistical and geometrical behavior of the uncontrolled parameters of a human face, including both the rigid transform caused by a head pose and the non-rigid transform caused by a facial expression. We focus on developing a 3D facial recognition scheme which is robust for these uncontrolled parameters. This thesis presents a novel 3D face recognition method by means of the evolution of iso-geodesic distance curves. Specifically, the proposed method compares two neighboring iso-geodesic distance curves, and formalizes the evolution between them as a one-dimensional function, named evolution angle function, which is Euclidean invariant. The novelty of this paper consists in formalizing a 3D face by an evolution angle functions, and in computing the distance between two faces by that of two functions. Experiments on Face Recognition Grand Challenge (FRGC) ver2.0 shows that our approach works very well on the neutral faces. By introducing a weight function, we also show a promising result on a non-neutral face database. A novel 3D surface segmentation scheme is developed to detect the partial similarity between two 3D facial images. The proposed algorithm is based on the iterative closest point (ICP) algorithm, which uses the mean square distance as the cost function and is not able to detect partial similarities. The presented thesis make an improvement of the ICP algorithm by iteratively removing points contributing the largest error, and the remaining area of surface can be shown to be the partial similarity between two surfaces. 3D Face Recognition From Range Image