摘要
针对车载LiDAR点云进行地面点滤波时,基于常规TIN、坡度等滤波算法不能根据局部地形变化自动调整阈值的问题,该文结合城市点云特征和地形起伏度,提出地形自适应的车载LiDAR点云滤波方法。该方法通过引入地形自适应参数进行区域增长阈值的动态调整,实现地面点、非地面点的自动精确滤波。通过实测数据试验,结果表明该方法可适用于车载LiDAR城市点云中地面点和非地面点的较精确分类,解决低矮浅丘、低矮灌木等地物点不容易正确分类的问题。
It is difficult to filter ground and non-ground points because of massive information, high density and three-dimensional discrete distribution of urban scene point cloud from mobile LiDAR, especially in the region of low hills and low shrubs. Based on the features of urban scene point clouds and relief, this paper proposed an au- tomatic adaptive terrain based filter method for the mobile LiDAR point cloud. The proposed method could adjust regional growth threshold automatically by adaptive terrain parameters, and realize the relatively more precise filte- ring of ground and non-ground points. An experiment was conducted with real LiDAR data, and the results showed that the propose method was suitable for urban scene point clouds from mobile LiDAR, and could solve the prob- lem of classification of low hills, low shrubs, etc..
出处
《测绘科学》
CSCD
北大核心
2015年第10期138-141,152,共5页
Science of Surveying and Mapping
基金
教育部创新团队资助项目(IRT13092)
中央高校基本科研业务费专项资金科技创新项目(2682013CX008)
