Real time face pose estimation using geometrical features

3D Head pose estimation using a hybrid pose estimation method is discussed in this paper. The only equipment used is simple webcam that is available on most computers and laptops today. The hybrid method consists of tracking facial landmarks of face and using geometrical face pose estimation to compute distances to estimate 3D position of head. The pose estimation system works real time as it is ultimately will be used for 2D-3D face recognition system. Actual data show reasonable error for rotation along each axis (Yaw, Pitch and Roll) by using only few facial landmarks.