基于双目视觉的SLAM四旋翼无人机  

SLAM Quadrotor UAV Based on Binocular Vision

作　　者：谭凯元 朱嘉林 邓君[1] 王荔蔷 杨家源 Tan Kaiyuan;Zhu Jialin;Deng Jun;Wang Liqiang;Yang Jiayuan(School of Mechanical Engineering,Dongguan Institute of Technology,Dongguan,Guangdong 523000,China)

机构地区：[1]东莞理工学院机械工程学院,广东东莞523000

出　　处：《机电工程技术》2022年第9期83-87,141,共6页Mechanical & Electrical Engineering Technology

摘　　要：为了实现在未知的环境下对周边的环境进行实时地构建地图,并且完成自主定位导航,设计了一种基于双目视觉的SLAM四旋翼无人机。在四旋翼无人机的基础上配置了双目摄像头,树莓派4B主控板,STM32驱动板及各种传感器,以ROS系统为载体,将ORB-SLAM建图,定位避障,自主导航各种算法集成在一起,对双目SLAM无人机进行设计与开发。无人机上装载的树莓派将双目摄像头采集到的信息进行筛选处理后,转化成点云图再输出,同时树莓派还要进行与飞控板之间的信息交流,实时控制无人机位移。最终将地图完整构建之后,再通过不同的算法规划和筛选出无人机从起点到终点的最优路径解,从而实现双目SLAM无人机的自主导航功能。实验结果表明,所设计的双目SLAM无人机能够构建出可靠有效的地图,并实现路径规划,自主导航。In order to realize real-time map construction of surrounding environment in unknown environment and complete autonomous positioning and navigation,a SLAM quadrotor UAV based on binocular vision was designed.On the basis of the quadrotor UAV,binocular camera,raspberry PI 4B master control board,STM32 driver board and various sensors were configured.With ROS system as the carrier,various algorithms of ORB-SLAM mapping,location obstacle avoidance and autonomous navigation were integrated together to design and develop binocular SLAM UAV.The raspberry PI loaded on the UAV filters the information collected by the binocular camera and converts it into a point cloud image for output.Meanwhile,the raspberry PI also communicates with the flight control board to control the displacement of the UAV in real time.Finally,after the complete construction of the map,different algorithms were used to plan and screen the optimal path solution of the UAV from the starting point to the end point,so as to realize the autonomous navigation function of binocular SLAM UAV.The experimental results show that the binocular SLAM UAV designed can construct a reliable and effective map,and realize path planning and autonomous navigation.

关 键 词：ORB-SLAM 四旋翼无人机 双目摄像头 地图构建 路径规划 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] V249[自动化与计算机技术—计算机科学与技术]