A Real-Time Hybrid Method for Head Pose Estimation
Kun Xu, Yufen Sun, Gang Liu · 2011
In this paper, we propose a hybrid method for head pose estimation which combines flexible method and tracking method. Firstly, flexible method detectes the frontal face of a driver and locates the position of feature points. Then the optical flow method is used to track these points, and head pose is calculated according to their tracking position. Our method has been tested with a head pose database. The results show it can perform well when the rotation angle of head is small.