B-Human 2021 – Playing Soccer Out of the Box

Thomas Röfer, Tim Laue, Arne Hasselbring, Lukas Malte Monnerjahn, Nele Matschull, Lukas Plecher

In: Rachid Alami, Joydeep Biswas, Maya Cakmak, Oliver Obst (Hrsg.). RoboCup 2021: Robot World Cup XXIV. RoboCup International Symposium (RoboCup-2021) 24th June 22-28 Virtual Seiten 302-313 Lecture Notes in Artificial Intelligence (LNAI) 13132 Springer 2022.


In 2021, our team B-Human participated in two events in the RoboCup Standard Platform League, namely the German Open Replacement Event and the actual RoboCup. We won both competitions. In both events, the biggest scientific challenge was to be able to play soccer on foreign robots in remote locations with the human team members neither being on site nor having direct access to these robots. We present our approach to a robot soccer system that only requires a fully automatic pre-game extrinsic camera calibration, but otherwise works out of the box. This paper focuses on the automatic calibration and some aspects of our lighting-independent computer vision system.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence