基于众源数据的自动化制图方法
An automated mapping method based on crowdsourced data
刘银 1伍伟绩 1王闯 1肖慧 1张京川1
作者信息
- 1. 河北全道科技有限公司北京分公司,北京 100102
- 折叠
摘要
为解决在传统高精度地图生产中,由于专业测绘采集车辆采集、制图手段的局限性,高精地图制作成本高、周期长、更新不及时等问题,本文采用了一种基于众源车辆进行自动制图的方法.将车端的环境感知结果、惯性导航系统(INS)/全球卫星导航系统(GNSS)组合导航数据等回传,在云端完成三维(3D)数据重建、矢量拓扑构建、成果质检发布等处理,从而完成自动化制图.选取某区域300 km的多场景路段进行实例验证,结果表明,本文所设计的技术方法能够解决高精地图鲜度问题及更新成本问题,可满足快速更新智能驾驶地图的要求,为智能驾驶的落地应用提供地图数据支撑.
Abstract
Traditional high-precision map production faces limited collection and mapping methods of professional surveying and mapping vehicles,resulting in high production costs,long cycles,and untimely updates of high-precision maps.To address these issues,this paper proposed a method for automated mapping based on crowdsourced vehicles.The environmental perception results and inertial navigation system(INS)/global navigation satellite system(GNSS)integrated navigation data of the vehicle were transmitted back,so as to complete the three-dimensional(3D)data reconstruction,vector topology construction,and result quality inspection and release in the cloud,thus achieving automated mapping.A 300 km multi-scenario road section in a certain area was selected for verification.The results show that the technical method designed in this paper can solve the problems of high-precision map freshness and update cost,meet the requirements of quickly updating intelligent driving maps,and provide map data support for the actual application of intelligent driving.
关键词
智能驾驶/自动化制图/众源数据/三维数据重建/矢量拓扑构建Key words
intelligent driving/automated mapping/crowdsourced data/three-dimensional(3D)/data reconstruction/vector topology construction引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据