Quantitative Comparison of Color Systems for Robot Soccer Applications

Robot soccer has evolved into a very dynamic and competitive field within the last few years, many different robot soccer leagues now exist. Most of the leagues rely on computer vision in one form or another to gather information about the game situation – the position of a team's robots, the position of the opponent's robots and the ball. While some problems robot soccer vision posed have been solved in the last few years, many still exist. One of the most notable current problems is robot vision systems capable of coping with (potentially sudden) lighting changes. In this work, we will deal with one aspect of all vision systems that has a major impact on their performance: the choice of color system. This work aims at a quantitative comparison of color systems with focus on applications in robot soccer vision. We will evaluate the power of the color systems in question regarding color recognition and color discrimination as well as their behavior in changing lighting conditions.