RoboCup中型机器人视觉系统的设计与实现  

Design and Realization of the Vision System for RoboCup Middle-Sized Robots

作　　者：行军[1] 杜边境[1] 李富强[1] 殷春浩[1] 

机构地区：[1]中国矿业大学理学院,江苏徐州221008

出　　处：《徐州工程学院学报（自然科学版）》2009年第2期50-55,共6页Journal of Xuzhou Institute of Technology(Natural Sciences Edition)

基　　金：教育部留学回国人员实验室建设科研基金资助项目(2003.18)

摘　　要：针对RoboCup中型机器人比赛,设计了机器人的视觉系统.首先选择HSI颜色空间并采用查找表技术对图像进行颜色分割,接着利用RLC技术分割感兴趣的区域,实现目标识别后,再采用角度定位和特征点定位相结合的方法实现机器人场上自定位.This paper focuses on the design of vision system with Robocup middle-sized soccer robot. The paper first chooses appropriate omni-directional mirror and a digital camera to set up a panoramic vision system. A method of pixel classification with lookup tables, which is based on HSI color model, is employed to segment the image. Then, RLC is applied to conduct the region segmentation. A localization model is established based on the combination of points-localization and angle-localization.

关 键 词：足球机器人 全景视觉系统 图像分割 机器人自定位 

分 类 号：TP242.62[自动化与计算机技术—检测技术与自动化装置]