Obstacle Avoidance and Dribbling Strategy for Humanoid Soccer Robots

This paper mainly investigates the obstacle avoidance and dribbling strategy for humanoid soccer robot by using only monocular vision, where the local object is used for positioning and recognizing the obstacles. Then by the relation between the object and the robot, the position of the robot at the world coordinate could be figured out with the ameliorative particle filter. It would be more robust and perform well for the obstacle avoidance and dribbling strategy to apply the above information. An enhancement scheme of the particle filter is also proposed in the paper. Furthermore, a real-time control strategy is developed for dribbling ball and avoiding obstacles. Finally, experiments verify the accuracy and robustness of the local positioning system, and show the feasibility and validity of the proposed obstacle avoidance and dribbling strategy for humanoid soccer robots.