基于点云数据的曲面重建算法比较研究  

Comparison of Surface Reconstruction Algorithms Based on Point Cloud Data

作　　者：吴旭 卢凌雯 梁栋栋 汪晓楚[2,3] WU Xu;LU Ling-wen;LIANG Dong-dong;WANG Xiao-chu(School of Computer and Information,Anhui Normal University,Wuhu 241003,China;School of Geography and Tourism,Anhui Normal University,Wuhu 241003,China;Geographic Big Data Research Center,Anhui Normal University,Wuhu 241003,China;Suzhou Surveying and Mapping Co,Ltd,Suzhou 221008,China)

机构地区：[1]安徽师范大学计算机与信息学院,安徽芜湖241003 [2]安徽师范大学地理与旅游学院,安徽芜湖241003 [3]安徽师范大学地理大数据研究中心,安徽芜湖241003 [4]苏州市测绘院有限责任公司,江苏苏州221008

出　　处：《安徽师范大学学报（自然科学版）》2019年第1期46-50,共5页Journal of Anhui Normal University(Natural Science)

基　　金：安徽省智慧城市与地理国情监测重点实验室2016年度开放性基金重点课题

摘　　要：针对现有的多种点云数据曲面重建算法,从曲面重建的网格曲面、隐式曲面、参数曲面三种不同重建方式入手,比较了四种算法针对不同目标物重建的优劣,并给出相应的精度评价。实验结果表明:基于NURBS参数曲面重建的方式最佳,基于贪婪投影三角化网格曲面重建的方式其次,基于移动立方体与基于泊松方程隐式曲面重建方式的时间复杂度与空间复杂度较大,且重建后的点云模型误差也较大。Aiming at the existing algorithms for surface reconstruction of point cloud data,this paper starts from three different reconstruction methods of surface reconstruction,such as mesh surface,implicit surface and parametric surface.The advantages and disadvantages of four algorithms for different target reconstruction are compared.And this paper gives the corresponding accuracy evaluation.The experimental results show that the method based on NURBS parametric surface reconstruction is the best,and the method based on greedy projection triangulation mesh surface reconstruction is the second,the time complexity and space complexity based on moving cube and implicit surface reconstruction method based on Poisson equation are larger,and the reconstructed point cloud model error is also larger.

关 键 词：贪婪三角化 移动立方体 泊松方程 NURBS 曲面重建 

分 类 号：O175.14[理学—数学]