A method for nose tip location and head pose estimation in 3D face data

3D nose tip location and head pose estimation are the crucial step for registration and recognition of 3D face data. In this paper, we propose an automatic method that is robust to head pose and expression variation. Firstly, we determine candidate areas using the curvature information. Then we define an effective statistical feature per point to extract the nose tip, in addition, we also estimate the head pose. Experiments are performed on the CASIA-3D face database which contains 123 subjects under varying pose and expression. Experiment results show that the method is capable of nose tip location and head pose estimation across the pose and expression variation represented in the database.