Soccer Field Boundary Detection Using Convolutional Neural Networks

Arne Hasselbring, Andreas Baude

In: Rachid Alami, Joydeep Biswas, Maya Cakmak, Oliver Obst (editor). RoboCup 2021: Robot World Cup XXIV. RoboCup International Symposium (RoboCup-2021) 24th June 22-28 Virtual Pages 202-213 Lecture Notes in Artificial Intelligence (LNAI) 13132 Springer 2022.


Detecting the field boundary is often one of the first steps in the vision pipeline of soccer robots. Conventional methods make use of a (possibly adaptive) green classifier, selection of boundary points and possibly model fitting. We present an approach to predict the coordinates of the field boundary column-wise in the image using a convolutional neural network. This is combined with a method to let the network predict the uncertainty of its output, which allows to fit a line model in which columns are weighted according to the network’s confidence. Experiments show that the resulting models are accurate enough in different lighting conditions as well as real-time capable.

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz