基于局部约束的建筑物屋顶点云平面分割方法

建筑物屋顶面点云分割结果的好坏对建筑物三维模型重建起着重要的作用。针对传统RANSAC算法建筑物屋顶面点云的分割问题,提出了一种基于局部约束的建筑物点云平面分割方法。利用点云局部曲面法向约束构建法向准则,利用半径约束的点云空间聚类的方法对共面屋顶面点云进行分解,从而抑制"伪屋顶面"的产生;利用局部抽样策略降低算法的迭代次数,减少运算量。实验表明该方法能够获得稳定可靠的建筑物屋顶点云分割结果,将有利于后续的建筑物三维模型重建。

A Building Roof Segmentation Method From Point Cloud Based on Local Constraints

The segmentation results of building roof point cloud play an important role in the reconstruction of 3D building reconstruction.Aiming at the problems of traditional RANSAC algorithm in the building roof point cloud segmentation,a plane segmentation method based on local constraints is proposed in this paper.In order to suppress the generation of "false roof plane",local surface normal constraint is used to construct the normal criterion and coplanar roof point cloud is broken down by a radius constrainted method of point cloud spatial clustering.A local sampling strategy is used to reduce the number of iterations of the algorithm.Experiments show that this method can obtain the stable and reliable results of the building roof point cloud segmentation,and will be beneficial to the subsequent 3D building model reconstruction.