Andre Treptow, Andreas Masselli and Andreas Zell
Real-Time Object Tracking for Soccer-Robots without Color Information
Proceedings of the European Conference on Mobile Robotics (ECMR 2003)
Objects in the RoboCup scenario (soccer playing robots) are identified by their unique color. These visual cues will be removed in the near future so that new vision algorithms are needed to cope with this.
In this paper we present a method for detecting and tracking the ball in a RoboCup scenario without the need of color information.
We use haar-like features trained by an adaboost algorithm to get a colorless representation of the ball. Tracking is performed by a particle filter.
It is shown that our algorithm is able to track the ball in real-time with 30 fps even in a cluttered environment.