Head movements degrade the image quality of high resolution Positron Emission Tomography (PET) brain studies through blurring and artifacts. Manny image reconstruction methods allows for motion correction if the head position is tracked continuously during the study. Our method for motion tracking is a structured light scanner placed just above the patient tunnel on the High Resolution Research Tomograph (HRRT, Siemens). It continuously registers point clouds of a part of the patient's face. The relative motion is estimated as the rigid transformation between frames. A geometric calibration between the HRRT scanner and the tracking system is needed in order to reposition the PET listmode data or image frames in the HRRT scanner coordinate system. This paper presents a method where obtained transmission scan data is segmented in order to create a point cloud of the patient's head. The point clouds from both systems can then be aligned to each other using the Iterative Closest Point (ICP) algorithm.