Abstract: In FIRA MiroSot robot soccer game，the vision system is the unique only way to get the position information of the robots and the ball on the playground.The speed and precision of the recognition of the vision system directly affect the victory or defeat of the game.The positions of all the entities obtained from the traditional vision system in the robot soccer game are inaccurate.According to the disadvantage，a new method of processing the real-time images using erosion and dilation algorithm in mathematical morphology to improve the precision of the recognition of the vision system is proposed in this paper.Experiment results show that the method can improve the precise of recognition on the premise of without reducing the speed of the recognition of the vision system in the game.

Key words: soccer robot, vision system, mathematical morphology, erosion, dilation

摘要： 在FIRA MiroSot机器人足球比赛中，视觉系统是获得比赛场上机器人与球位置信息的唯一途径。视觉系统的识别速度、精度直接影响到比赛的胜负。针对传统视觉系统在机器人足球比赛中获取各实体的位置不够准确的问题，提出了一种结合数学形态学中腐蚀/膨胀算法来处理视觉系统中的实时图像，增加足球机器人视觉系统识别的精度的设计方案。实验结果表明，该方案在没有降低比赛中识别速度的前提下，大大地提高了识别精度。

关键词: 足球机器人, 视觉系统, 数学形态学, 腐蚀, 膨胀