基于车载激光扫描的城市典型道路面点云提取  

Point cloud extraction of typical urban road surface based on vehicle mounted laser scanning

作　　者：杜珍珍 赵海洋 DU Zhenzhen;ZHAO Haiyang(Zhejiang Provincial Land Survey and Planning Company Limited,Hangzhou Zhejiang 310030,China)

机构地区：[1]浙江省国土勘测规划有限公司,浙江杭州310030

出　　处：《北京测绘》2023年第2期260-264,共5页Beijing Surveying and Mapping

摘　　要：针对城市典型道路结构特征,提出一种车载道路面点云数据提取方法。根据高程阈值算法对原始车载激光扫描点云数据进行滤波处理,得到地面点,再使用改进区域生长算法对地面点进行处理,提取得到道路面点云。通过实测车载激光扫描数据进行实验,结果表明,本文提出方法提取道路面点云结果的检测质量q、完整性r以及准确性p均在93%以上,本文方法是行之有效的。According to the structural characteristics of typical urban roads,a method for extracting point cloud data of vehicle road surface is proposed.According to the elevation threshold algorithm,the original vehicle borne laser scanning point cloud data is filtered to obtain the ground points,and then the improved region growth algorithm is used to process the ground points to extract the road surface point cloud.The experimental results show that the detection quality q,integrity r and accuracy p of the proposed method for extracting road surface point cloud results are all above 93%,and the proposed method is effective.

关 键 词：车载激光扫描 高程阈值滤波 曲率值 法向量 区域生长算法 

分 类 号：P234[天文地球—摄影测量与遥感]