The purpose of this paper is to design and implement a robot soccer game to conform RoboCup middle size league. First, according to the rules of RoboCup, we design the middle size soccer robot, which includes the independent image processing system and mechanical architecture. And the soccer robot equips the on-board computer system and interface circuits to make decisions. Second, the omnidirectional vision sensor of the vision system deals with the image processing and positioning for obstacle avoidance and target tracking. The boundary-following algorithm (BFA) Is applied to find the important features of the field. In the control strategies, we present five state modes, which include the chasing ball mode, the catching ball mode, the shooting ball mode, obstacle avoidance mode and the defense mode. The soccer robot implemented for testing the performance of the proposed algorithm has shown that it chases a dynamic target successfully in the competition. And the experimental results confirm the methods finally.