A real-time stereo head pose tracking system

Abstract

Head pose tracking is very important in many human machine interfaces. In this paper, a stereo head pose tracking system is proposed, which detects and tracks six facial features (pupils, nostrils and lip corners) automatically when a human face appears in front of two low-cost USB cameras. The epipolar constraint and a 3D facial feature model are used to improve the accuracy and robustness of the system. This system has the advantages of automatically detecting the facial features and recovering the features lost during the tracking process. Encouraging results have been obtained using the proposed system