Robust pose estimation using Time-of-Flight imaging

We propose a computer vision algorithm for head pose estimation that is suitable for real time environments. Our solution is based on the depth information provided by a Time of Flight camera, image processing algorithms for facial landmarks detection and support vector machine classification. We study the effect of different factors including head pose angles, background reflective surfaces and computing duration. We identify an extended range of roll, yaw and pitch rotations angles for which the algorithm provides reliable estimates.