摘要
本文采用基于决策树的人工智能方式,通过对点云数据分割、提取分割片轮廓,手动设置地物的约束条件来提取LiDAR点云中的地物。两组实验表明,本文的方法能够成功提取点云数据中的建筑物、车辆等地物;要提高精度,需增加点云密度,因此本文方法适用于较平坦的城镇地区。
This paper adapted the method of artificial intelligence which is based on decision tree to extract features from LiDAR data. The main steps included segmenting, deriving boundary and extracting features by setting parameters manually. In the end, the results of two tests showed that the method could extract building and cars from laser data effectively.
出处
《测绘科学》
CSCD
北大核心
2013年第3期118-120,共3页
Science of Surveying and Mapping
关键词
点云加密
分割片
轮廓线
知识
地物提取
densification of point clouds
segment
outline
knowledge
feature extraction
