Building Roof Extraction from LIDAR Point Clouds Using Clustering Techniques
This paper presents a method for buildings roof extraction from LIDAR points cloud. Firstly, some possible roofs boundary points were extracted by constructing the TIN model of the buildings points cloud; secondly, points located on different slope roofs planes are segmented by clustering techniques on directions of points normal vectors: and lastly, a distance clustering algorithm was used to find boundary points between stagger roofs. After those three steps, a probably borderline of a building roof can be calculated by connecting its boundary points in turn. In order to get a regularization borderline, a boundary expanding algorithm is presented at last. And the experimental result shows that this method is able to extract effectively all kinds of polygon building roof.
LIDAR points cloud building roofs extraction clustering TIN
MAO Jianhua ZENG Qihong LIU Xuefeng HUANG Rui
School of Communication and Information Engineering, Shanghai University, Shanghai 200072, China
国际会议
武汉
英文
297-304
2011-06-26（万方平台首次上网日期，不代表论文的发表时间）