The Research On The Robot-soccer Visual Subsystem

In 1992 , MACK WORTH put forward the concept of that develop a team of fully autonomous humanoid robots that can win against the human world soccer champion team by the year 2050.The project is to promote AI, robotics, and related field, it is an attempt to foster AI and intelligent robotics research by providing a standard problem where wide rang of technologies can be integrated and examined. FIRA chose to use soccer game as a central topic of research, aiming at innovations to be applied for socially significant problems and industries. In order for a robot team to actually perform a soccer game, various technologies must be incorporated including: design principles of autonomous agents, multi-agent collaboration, strategy acquisition, real-time reasoning, robotics, and sensor-fusion. FIRA is a task for a team of multiple fast-moving robots under a dynamic environment. One of the major applications of FIRA technologies is a search and rescue in large scale disaster. The project specially research in social significant issues. There is no doubt that the soccer robot is an emerging interdisciplinary study. Its technology mainly includes 4 subsystems: Robot subsystem, visual subsystem, wireless communication subsystem and policy-making subsystem.Visual system is the important apparatus for human being and people acquire most of information from the outside world by their vision systems. In the process of human being's exploration to the nature world, man always hope to find some machine-computing ways to achieve human being's vision function, and to acquire information automatically from outside world.The visual subsystem as the soccer robot main sensation equipment is the entire robot eye, which is the most main channel that gains the extraneous information and the foundation of soccer robot assembly system. It continuous gathers on the information of the field and transmits the result of processing the picture to the decision system.The quality of visual system decides the robot performance display. as a result of the soccer robot quick movement and the intense resistance, the video frequency is 30/s, the computer must complete visual information processing,the communication,the action during 33.34ms and forms the next cyclical competition strategy. This request the visual system is characteristic of real-time. robustness and accuracy. Therefore, the article put emphasis on the kind of robot visual subsystem.The research contents include:(1) Research in the structure, function and software design of all Morisot subsystems.(2) A new camera calibration method of varying effective focal length by neural network model is presented.(3) It discuss the principles and methods of color perception and image segmentation based neural network, and analyzes the demands and characters of robot-soccer visual system, finally presents the basic principles for image processing. A new method of correct image based on neural network is presented and the image is segmented.(4) The HSI color space was applied to set up color partition and color database to implement the partition of object images. A more reasonable and anti-jamming design was put forward by comparing and analyzing various representative color tag designs with a fast and relatively exact successive approximation algorithm given correspondingly for angular compensation. After analyzing the demands and characters of robot soccer vision system, it offers a quick method of object recognition based dynamic windows.(5) A new method of obtaining the central of ball, which is based on the model of distortion difference and the space analytic geometry theory, is presented because the ball central point obtains difficulty so that we often lose the object of robot or enlarge the difference of the path of robot.(6) The visual subsystem based on image processing,neural network,visual C++ is developed and applied to competition that can adapt to the real-time of competition.Robot soccer match system is involved in many technology domains which have a wonderful application foreground and make it possible for artificial intelligence researchers and robot research to communicate with each other, which will accelerate the development of related technologies, and have a good effort of science and technology popularization.