基于代数曲面的残缺点云数据迭代修复算法  

Iterative Repair Algorithm for Incomplete Point Cloud Data Based on Algebraic Point Set Surface

作　　者：胡少乾 周来水[1] 闫杰琼 文思扬 穆冬梅 HU Shaoqian;ZHOU Laishui;YAN Jieqiong;WEN Siyang;MU Dongmei(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]南京航空航天大学机电学院,江苏南京210016

出　　处：《机械制造与自动化》2022年第2期83-86,共4页Machine Building & Automation

基　　金：国家自然科学基金项目(52075258)。

摘　　要：针对逆向工程领域,激光扫描仪获取的散乱点云数据难免存在着残缺的问题,设计一种迭代修复残缺点云孔洞的算法,对其进行下采样,基于加权局部最优投影算子得到分布均匀的稀疏点集;将点沿稀疏点集的代数点集曲面排斥推向残缺区域;对稀疏点集进行上采样完成孔洞区域的一次修复。将上述步骤迭代应用,逐渐缩小残缺区域,直到完成修复。研究表明:该方法对表面曲率连续的残缺点云模型可自适应完成修复,恢复原有残缺区域的几何信息。To address the incomplete scattered point cloud data obtained by laser scanner in the field of reverse engineering, an iterative algorithm to repair the holes in the residual cloud is designed. The sparse point set is downsampled, and the uniformly distributed sparse point set is obtained based on the weighted local optimal projection operator. The points are pushed to the incomplete area along the algebraic point set surface of the sparse point set. The sparse point set is upsampled to complete the repair of the hole area. The above steps are applied iteratively to reduce the incomplete area gradually until the repair is completed. The research shows that the method can adaptively repair the incomplete point cloud model with continuous surface curvature, and recover the geometric information of the original incomplete area.

关 键 词：残缺点云数据 点云修复 代数点集曲面 点云上采样 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]