Building Plane Segmentation Method Based on Dense Image Matching Point Cloud Data
Aimed at the problem that current point cloud segmentation methods are easy to be over-segmented and under-segmented in Dense Image Matching(DIM)point cloud,a building plane segmentation method based on DIM point cloud data is proposed in the paper.Firstly,the original point cloud is segmented by local elevation and projected to xoy plane.Secondly,noise is removed according to the density distribution of point cloud,and the building contour line segment is extracted by RANSAC combined with Euclidean clustering algorithm.Then,the extracted line segment is used as constraint condition to segment the building elevation point cloud.Finally,the roof point cloud is extracted by using European clustering algorithm.Experimental results show that this method can effectively segment DIM building plane point cloud.
building point cloudDIMpoint cloud segmentationconstraint conditionRANSAC
万方数据
维普
