This bachelor thesis is for analyzing possibilities of integrating the face tracking software FaceAPI (Seeing Machines, Braddon, Canberra, Australia) in medical treatments like the robot based transcranial magnetic stimulation (TMS). A common use of the TMS is the therapy of tinnitus. The FaceAPI offers to track a human face and to get the head position regarding the basis coordinate frame of the used webcam. An integration of the FaceAPI would increase safety and comfort for patients during the robot based TMS. For this purpose we evaluate the accuracy of the FaceAPI by comparing with the accuracy of the Polaris Spectra tracking system (Northern Digital Inc., Waterloo, Ontario, Canada). The Polaris Spectra system is actually used for navigation in the robot-based TMS and could be supported or even replaced by the FaceAPI. For realizing the comparison we track a model of a human head which is moved by an Adept Viper s850 industrial robot (Adept Technology Inc., Livermore, California, USA) and evaluate the recorded data. For a better understanding theoretical backgrounds of controlling a robot using data recorded by head tracking systems are presented. To facilitate the exchange of other tracking systems and to offer an easy and comfortable use via TCP/IP interface the FaceAPI will be integrated in the tracking system developed by the Institute for Robotics and Cognitive Systems.
|Bachelor of Science
|Published - 01.09.2009