自主足球机器人的单目视觉自定位方法  

A Self-Localization Method with Monocular Vision for Autonomous Soccer Robot

作　　者：谢云[1] 杨宜民[1] 

机构地区：[1]广东工业大学自动化学院,广东广州510090

出　　处：《微电子学与计算机》2005年第10期129-132,共4页Microelectronics & Computer

基　　金：国家863计划项目(512-9935-08);广东省重点科技攻关项目(200027)

摘　　要：全自主机器人足球系统中,每个自主足球机器人都是一个独立的智能机器人,有自己的决策、通信、视觉和感知、机电控制等子系统,但是系统不具备全局视觉。为了使系统能够了解场上全局信息,足球机器人的自定位是必须解决的关键问题之一。文章介绍了一种在颜色场地信息已经视觉识别的前提下,通过单目视觉测距,实现全自主足球机器人自定位的几何方法;对摄像机CCD像面的量化、镜头的非线性畸变和摄像机定标等可能引起误差的因素进行了分析,实验结果表明方法可行。In autonomous robot soccer system, every autonomous soccer robot is an unattached intellective robot. The robot has its decision subsystem, communication subsystem, vision and multi-sensor subsystem, machine-electric control subsystem etc, but-the system hasn＇t any overall vision. To get the overall information of the field, the self- localization of autonomous soccer robot is one of sticking points and must be solved. In this paper, a geometrical self-localization method of autonomous soccer robot is introduced. It is a precondition that some color landmarks in the field have been recognized by cyber-vision. A method of depth measurement by monocular vision is used in the geometrical self-localization method. And three possible factors which bring ranging errors are analyzed. They are quantization error of CCD image element, no-linearity aberration of its camera lens, and vidicon calibration etc. It＇s shown that the geometrical self-localization method is feasible.

关 键 词：全自主足球机器人 单目视觉 测距 几何定位 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]