基于区域投影的微型足球机器人系统识别  

Vision Recognition in Micro-Robot Soccer System Based on Region Projection

作　　者：张艳珍[1] 袁野[1] 巩玉国[1] 欧宗瑛[1] 王金鹤[2] 

机构地区：[1]大连理工大学CAD&CG研究所,大连116023 [2]海信集团技术中心,青岛266071

出　　处：《机械科学与技术》2001年第2期318-320,共3页Mechanical Science and Technology for Aerospace Engineering

摘　　要：针对足球机器人这一多机器人系统中机器人小车和球的视觉识别 ,开发了一种基于区域投影的识别方法。首先对于由 CCD镜头所捕获的图像进行宏观位置跟踪 ,即自上而下、自左至右地进行网格搜索 ,获取目标物体 (机器人小车和小球 )内一点 ;然后从该点出发 ,利用区域投影 ,对目标物体精确定位 ,得到小球和各机器人的中心位置及各机器人的方位。运行结果表明 ,本方法具有较高的识别精度 ,能够很好地识别处理多个机器人粘连在一起的情况 ,同时识别速度较快 ,对于大小为 80 0 pixels6 0 0 pixels(包括六个机器人小车 )的图像 ,采用 PII35 0 PC识别速度达到每帧仅用时 1.2 m s。This paper deals with the vision recognition system for recognizing vehicle robots and football based on object projection in micro-robot soccer system. The vision recognition is processed with two stages. In the first stage, a grid search scanning from top to bottom and left to right is processing to find an object point inside an unrecognizing object (robots or football), then a precision analyzing and recognizing procedure based on object region projection is processing for determining the robots and football's position and also the robots' orientation in the second stage. The recognition algorithm is robust in treating noisy image captured by CCD, even in the situation while multiple vehicle robots are closely adjacent to each other. The algorithm is also very efficient, it takes only 1.2 ms in PII350 PC for recognizing and identifying six vehicle robots in a captured frame, which is composed of 800 pixels × 600 pixels.

关 键 词：视觉识别 足球机器人 区域投影 图像分析 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]