Vision Guided Soccer Robot

Humans carry out daily activities through synchronization of different body parts. Similarly, synchronization of sub-behaviors would allow the humanoid robot to play soccer as a player. A series of sub-behaviors are proposed and created which are then divided into two categories called Red Ball Detection and Goal Post Detection which utilizes the color blob statistics, trigonometric depth analysis, and motion analysis. Motion includes Rotate and Kick Motion which utilizes the motion analysis. A series of investigation would also be done on the trigonometric depth analysis and motion analysis to further enhance the sub-behavior that would be utilizing them. The quantitative results from analysis helps to execute the behavior efficiently. Nao humanoid robot have been used for the experimental studies.

I. Introduction (Heading 1)

The humanoid robots are created to mimic the activities of humans as much as possible. However, there are still limitations on behaviour building techniques and executions [12], [24]. Although some tasks might seem easy for humans to perform, it might consist of several synchronized movement for the humanoid robot to do. For instance, a soccer player will first need to detect the ball, move towards it, detect the goal, face the goal before kicking it and execute the task towards the goal. Few behaviors are proposed using vision and robot movements such as head-yaw movement, move-towards the ball, rotating-face to goal post and kick-motion.