Motion capture and detection is one of the main sections in our rapidly developing world of information. Due to the large scale development in IT there is a huge demand for such systems and the improvement thereof. Motion capture and detection systems are used in many industries such as medical sciences or the military industry. Nowadays the most significant area is the entertainment industry including the movie and video game industry. Due to the fast growth of computing capacity of embedded systems and thanks to the inexpensive Inertial Measurement Units - IMU, it has become possible to manufacture "smart" sensors in a cost-efficient way. The sensors in a networked structure are able to capture and transmit the motion of a human being in real time.
Apart from motion capture systems, humanoid robots, even if not life-sized, are becoming more and more accessible to the regular public. However the technological development is far from what has been imagined. These days, it is a big challenge for developers to control even primitive robots in a realistic way.
As part of my thesis, it was my goal to design and implement such a motion capture system which on one hand is based on inertial mesurement units and which on the other hand is able to capture and detect the motion of a human being with minimal delay in a realistic way. Last but not least it was another goal of mine to control a humanoid robot with these captured motion data.