基于双目视觉技术的复杂环境下机器人自动导航研究  

Research on robot automatic navigation in complex environment based on binocular vision technology

作　　者：吕强 张海涛 王辉 李永强 Lv Qiang;Zhang Haitao;Wang Hui;Li Yongqiang(Ultra High Voltage Company of State Grid Gansu Electric Power Company,Gansu Lanzhou,730070,China)

机构地区：[1]国网甘肃省电力公司超高压公司,甘肃兰州730070

出　　处：《机械设计与制造工程》2023年第9期79-84,共6页Machine Design and Manufacturing Engineering

摘　　要：为了规划机器人行驶路径,保障机器人行驶安全,提高机器人任务完成的安全性与自动性,提出基于双目视觉技术的复杂环境下机器人自动导航方法:标定双目摄像机,利用双目视觉技术采集机器人行驶环境图像;提取机器人行驶环境图像特征;按双目视觉三维重建原理,获取机器人行驶环境三维点云数据;以机器人行驶环境二维地图为依据,构建机器人行驶环境栅格地图;采用A*算法在构建的栅格地图中规划机器人行驶路径,完成机器人自动导航。实验表明:该方法可实现复杂环境下机器人自动导航,并规划出较为理想的机器人行驶路径。In order to plan the robot's driving path,ensure the robot's driving safety and improve the safety and automaticity of robot's task completion,an automatic navigation method of robot in complex environment based on binocular vision technology is proposed,the binocular cameras are calibrated,the images of robot's driving environment are collected by binocular vision technology.It extracts the image features of the drive environment of the robot.According to the principle of binocular vision three-dimensional reconstruction,the three-dimensional point cloud data of robot driving environment are obtained;Based on the two-dimensional map of robot driving environment,a grid map of robot driving environment is constructed;A*algorithm is used to plan the robot's driving path in the constructed grid map to complete the robot's automatic navigation.This method can realize robot automatic navigation in complex environment and plan an ideal robot driving path.

关 键 词：双目视觉技术 复杂环境 机器人自动导航 路径规划 A*算法 栅格法 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]