基于自发光标志物识别的室内机器人定位研究  

Study on positioning method of indoor robot based on feature recognition of self-luminous markers

作　　者：石晓华 王国庆[1] 程汉 刘平利 王辉 王强[1] SHI Xiaohua;WANG Guoqing;CHENG Han;LIU Pingli;WANG Hui;WANG Qiang(School of Construction Machinery,Chang’an University,Shaanxi Xi’an 710064,China)

机构地区：[1]长安大学工程机械学院,陕西西安710064

出　　处：《工业仪表与自动化装置》2022年第3期60-63,共4页Industrial Instrumentation & Automation

基　　金：陕西省重点研发计划(2021GY-039)。

摘　　要：针对室内光线条件不良或者特征物不明显造成机器人的视觉定位与导航困难的问题,提出了一种基于自发光标志物的特征识别定位方法,实现了机器人室内位置的追踪。首先通过放置在屋顶的相机来获取机器人的整个工作环境,对机器人的工作环境进行整体定位,再使用SURF算法进行特征提取与匹配,根据图像处理结果确定机器人的具体位置,进行机器人位置的实时跟踪。实验结果表明,该方法在精确度要求不高的情况下能够获取到机器人的位姿信息,证明了整个理论方案的可行性。Aiming at the difficulty of visual positioning and navigation of robot caused by poor indoor light conditions or unobvious features,this paper proposes a positioning method based on feature recognition of self-luminous markers,which realizes the indoor position tracking of the robot.Firstly,the whole working environment of the robot is obtained by the camera placed on the roof,and the overall positioning of the working environmentis carried out.Then,SURF algorithm is used for feature extraction and matching.The specific position of the robot is determined according to the processing results of the image,and the real-time tracking of the robot position is carried out.The experimental results show that the method can obtain the pose information of the robot under the condition of low accuracy requirement,which proves the feasibility of the whole theoretical scheme.

关 键 词：自发光标志物 视觉定位 特征识别 室内机器人 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]