Robust 3D head mracking using camera pose estimation

In this paper we present a robust method to recover 3D position and orientation (pose) of a moving head using a single stationary camera. Head pose is recovered via a camera pose estimation formulation. 3D feature points (artificial or natural occurring) are acquired from the head prior to tracking and used as a model. Pose is estimated by solving a robust version of "Perspective n Point" problem. The proposed algorithm can handle self occlusions, outliers and recover from tracking failures. Results were validated by simultaneous tracking using our system and an accurate magnetic field 3D measuring device. Our contribution is a system that is not restricted to track only human heads, and is accurate enough to be used as a measuring device. To demonstrate the applicability of our method, three types of heads (human, barn owl, chameleon) were tracked in a series of biological experiments.