三维激光点云中建筑物立面边界特征提取方法

Boundary feature extraction method of building facade in 3D laser point cloud

提取建筑立面中窗户等非墙面元素的边界特征是利用点云数据进行建筑物精细化建模的难点。针对该问题,提出一种基于模板匹配的建筑物激光点云立面几何特征提取方法。首先采用RANSAC算法粗分割建筑物立面,通过双向DBSCAN算法提取墙面点云,并依据墙面点云提取墙面边界和分离窗户点云;其次,对窗户点云进行欧式聚类和矩形拟合分类;然后对窗户的特征凸包点内插获得窗户边界模板点云;最后利用模板匹配方法获取完整的建筑物立面窗户点云边界,并获得整体建筑物立面边界点云数据。实验结果表明,该方法可以精确提取建筑物墙面点云边界,补全部分边界缺失的窗户轮廓边界点云,满足建筑物精细化建模的需求。

To extract the boundary features of non-wall elements such as windows in building facade is a difficulty in building fine modeling by using point cloud data.To solve this problem,this paper proposed a geometric feature extraction method of building laser point cloud elevation based on template matching.Firstly,random sample consensus(RANSAC)algorithm was used to roughly segment the building facade and bi-directional density-based spatial clustering of applications with noise(DBSCAN)algorithm was used to extract the wall point cloud.The wall boundary and window point cloud were extracted according to the wall point cloud.Secondly,euclidean clustering and rectangle fitting were used to classify the window point cloud.Then the feature convex hull points of the window were interpolated to obtain the window boundary template point cloud.Finally,the template matching method was used to obtain the complete building facade window point cloud boundary and the whole building facade boundary point cloud data.The experimental results show that this method can accurately extract the boundary of point cloud between building wall and window,which can meet the needs of building fine modeling.