To Oriented Mirosot The Visual System Design Research,

The micro-robot soccer tournament (MiroSot) system is a typical centralized control model based on central vision and decision-making subsystem. For MIROSOT large-league, some new practical problem for the vision system needs to be solved. In this thesis, the technologies in designing such a new vision system are investigated.Firstly, we give a brief overview to the structure of the robot vision system . Technical requirement and important functions have been given to mobile robot vision system based on analyzing machine vision theory and vision guide techniques. And we put forward the project we adopt in our system.Secondly, we emphases on some key technologies of new vision system. To identify the targets robustly, we exploder the technology of color-forecasting based on HSV. The dynamic track and identification to the color-ball in soccer robot system based on vision is the most important task. In order to overcome the difficulty in identifying the ball sheltered by mobile robots in the robot soccer games, this paper presents a new visual tracking identification method based on an adaptive search window and the prediction to the location of the ball. The use of wide-angle lens causes geometrical distortion in actual images. In order to remove it, based on the compare of several typical ways a method is presented for meeting need of competition.Lastly, we discuss the design and implement of the new system. The subsystem has been tested, and the result shows a satisfactory performance in competitions.