Research Of Self-localization System Based On Omni-directional Vision On Soccer Robot

Robot soccer is a researchful hotspot of robot control, machine vision and artificial intelligence presently. The thesis designs and implements a novel Omni-vision system for the application in the middle-size league of RoboCup-The World Cup of Soccer Robots and for the platform of autonomous robot, which composed of a camera and a mirror generated with a surface of revolution, and does research on the robot's self-localization methods based on the Omni-vision system. The system can capture the rather perfect panoramic image real timely, and process and segmentation image, which made robot can get the working environment of itself exactly, and complete robot's self-localization and recognization of target from the extraction of features, to satisfy the requirement of real time and accuracy of compete.This paper works over the following contents: introduce several Omni-mirror in common use, and choose a suit of Omni-vision system which is the same with the middle-size league of RoboCup according diversified requirement of compete for vision system. Design a image processing algorithm for the panoramic image, which can segment the image by color, extract the features of image fast and effectively and complete the recognization of target, and then realize the robot's self-localization based on the character of the imaging. Put forward an extend kalman filter(EKF) arithmetic based on the analysis of robot sport framework and theory of panorama vision system, which amalgamation the data information from mile meter and vision system, and then complete robot's self-localization.Lastly, the thesis summarizes the whole paper, make out the main researchful fruit, at the same time point out the deficiency and places need to be improved.