摘要
基于车载点云数据开展道路要素的自动提取与分类是三维激光扫描技术服务于城市运维活动的一项重要应用。基于现阶段的行业研究情况,本文首先开展了针对不同类型道路要素对象进行方法及算法汇总,然后基于应用实例开展了车载激光雷达(LiDAR)点云用于道路结构、路面标识以及道路区域杆状地物等主要道路要素的自动提取与分类的应用,对其实用性和准确性进行了评价。
Automatic extraction and classification of road elements based on vehicle point cloud data is an important application of 3D laser scanning technology in urban operation and maintenance activities.Based on the current industry research situation,this paper firstly summarized a series of methods and algorithms for different types of road feature objects,and then carried out the application of vehicle-borne LiDAR point cloud for automatic extraction and classification of main road features such as road structure,road marking and rod-shaped features in road area based on application examples,and evaluated its practicability and accuracy.
作者
李冠
孟祥武
LI Guan;MENG Xiangwu(BGI Engineering Consultant Company Limited,Beijing 100038,China;Beijing Institute of Surveying and Mapping,Beijing 100038,China)
出处
《北京测绘》
2023年第7期969-974,共6页
Beijing Surveying and Mapping
基金
国家自然科学基金(42274029)。
关键词
车载激光雷达点云
道路要素
自动提取与分类
局部特征约束
聚类分析
语义分割
样本训练
vehicle light detection and ranging(LiDAR)point cloud
road elements
automatic extraction and classification
local feature constraint
cluster analysis
semantic segmentation
sample training
