一种基于点云数据的建筑物平面精细分割方法

针对现有大规模点云数据平面特征分割方法中存在的错误识别、效率低、抗噪性差等问题,该文提出一种基于2D霍夫变换和八叉树的建筑物平面精细分割方法。该方法首先,对原始点云进行空间均匀降采样并向X-Y面投影,利用改进的2D霍夫变换算法提取投影后的点云线段,使用选权迭代法精确计算线段所在直线的方程及端点坐标,进一步确定立面的空间几何方程;接下来,建立原始点云数据的八叉树结构,利用端点坐标设计立方体并分割出立方体内的立面点云;最后,将立面点云从原始点云中剔除,对余下点云降采样并向X-Z面投影,重复以上过程分割水平面点云。试验验证了该文方法对建筑物面状特征分割的有效性。

Fine segmentation of building planar feature from large scale unorganized point cloud

Aiming at the problems existing in the planar feature segmentation methods of large scale point cloud data,such as misidentification,low efficiency,and poor anti-noise,etc.,this paper presented a novel method for segmentation of planar feature from unorganized point cloud based on 2 D Hough transform and octree.Given the input point cloud,three steps were performed to segment planar feature.Firstly,the original point cloud was downsampled and projected to the X-Y plane,where an extended 2 D Hough transform algorithm was employed to extract the line segments,followed by selecting weight iteration to calculate line equations and endpoint coordinates of those line segments,and then we could determine the space geometric equations of the vertical planes.Secondly,the octree structure of the original point cloud was established,then we used the exact endpoint coordinates of the line segment to design a cube perpendicular to the X-Y plane and all points belonging to this cube were extracted.Finally,the facade point cloud was removed from the original point cloud,and remaining point cloud was downsampled and projected to the X-Zplane.The above process was repeated to segment the horizontal planes.Experiments results indicated that the proposed algorithm was suitable for segmentation of building planar feature in different scene.