摘要
针对机器人同步定位与导航系统中,空中无人机执行地面任务灵活性差、地面无人车视野易被遮挡等问题,采用基于视觉同步定位与地图构建(SLAM)的空地协同导航方法;针对现有空地协同导航系统中无人机多视角下的视差问题,提出了3D-2D线特征匹配方法;针对图像轮廓还原程度较差的问题,将无人机关键帧拼接后的边缘图像与SLAM地图进行基于轮廓相似度的融合。地面无人车通过视觉标靶对自身定位,并融合激光雷达局部地图的方法实时修正地图,完成路径规划。通过仿真进行测试,证明了方案的可行性和方法的有效性。
In the synchronous positioning and navigation system of robots,the flexibility of aerial unmanned aerial vehicles(UAVs)in performing ground tasks is poor,and the vision of ground unmanned aerial vehicles(UAVs)is easily blocked.In order to solve the parallax problem of uav under multi-view in the existing air-ground cooperative navigation system,a 3D-2D line feature matching method is proposed.Aiming at the problem of poor image contour restoration,the edge image of UAV keyframe Mosaic and SLAM map are fused based on contour similarity,and the global map constructed by this method has both accurate position information of the former and accurate shape information of the latter.The ground unmanned vehicle locates itself by visual target,and makes path planning in global map by two-way RRT algorithm.In the process of moving,the local map is constructed by lidar in real time,so as to avoid the new obstacles,update the global map,and re-plan the path until the target point is reached.The feasibility and effectiveness of the proposed method are proved by testing the proposed method in simulation environment.
作者
高宏伟
于斌
武炎明
GAO Hongwei;YU Bin;WU Yanming(Shenyang Ligong University,Shenyang 110159,China;Shenyang Institute of Automation,Shenyang 110017,China;不详)
出处
《沈阳理工大学学报》
CAS
2022年第1期7-13,共7页
Journal of Shenyang Ligong University
基金
辽宁省自然科学基金项目(2020-BS-026)
辽宁省教育厅高校创新人才项目(LR2019058)。
关键词
协同导航
视觉同步定位与地图构建
线特征
路径规划
collaborative navigation
visual simultaneous localization and mapping
line features
path planning
