一种基于散乱点云的边界提取算法

A BOUNDARY EXTRACTION ALGORITHM BASED ON SCATTERED POINT CLOUD

点云边界是曲面的重要特征之一,边界线的快速准确提取对于提高曲面重构的效率和质量具有重要意义。首先,采用基于kd-tree搜索的方法建立点云空间拓补关系,进行K邻域快速搜索,以采样点及其K邻域作为局部型面参考依据拟合微切平面,将其向微切平面投影;其次,在微切平面上建立局部坐标系,并对投影点进行参数化,根据邻域点集在采样点处的场力大小之和可以表示点集的平均作用来识别点云的边界特征点;最后,从提高边界线连续性的角度,利用NURBS曲线插值方法连接边界线。实验结果表明,该算法可以快速、有效地提取出点云的边界特征点,并得到C2连续的边界线,满足曲面重构的要求。

The boundary of point cloud is one of the most important features of the surface,extracting the boundary line quickly and accurately is important to improve the efficiency and quality of surface reconstruction.First,we use kd-tree based searching method to establish topological relationship in cloud point space and carry out K neighbourhood fast search,and fit the micro tangent plane by using the sampling points and its K neighbourhood as the reference basis of the partial type surface,then project these points onto the micro tangent plane.Secondly,we set up a local coordinate system on the plane and parameterise the projecting points,and identify the boundary characteristic points according to the theory that the sum of the field force magnitudes of the neighbourhood point sets on sampling points can represent the average effects of the point sets.Finally,in perspective of improving the continuity of boundary lines,we use NURBS curve interpolation method to connect the boundary lines.Experimental results show that the algorithm can extract the boundary features of point clouds quickly and effectively,and get the boundary line with C2 continuity,meets the requirement of surface reconstruction.