基于散乱点云的三维服装建模

3D Garment Modeling Based on Scattered Point Cloud

基于实体服装的散乱点云数据提出一种快速生成三维服装模型的方法。通过深度相机扫描设备实现服装表面三维点云数据的快速提取,并进行预处理;结合区域生长算法和Delaunay三角剖分算法,提出一种以无向点云为输入,通过三角剖分的形式生成插值曲面的重建算法对服装模型进行重建;将基于散乱点云的三维服装建模方法与贪婪投影三角化法和泊松法对比分析。结果表明:该方法重建的服装模型更好,得到的网格分布均匀,面片光滑,并且保留了服装点云原有的形状。

Based on the scattered point cloud data of solid garment,a method of quickly generating a 3D clothing model is proposed.The depth camera scanning device is used for rapid extraction the 3D point cloud data of the garment surface and perform pre-processing.Combining region growing algorithm and Delaunay triangulation algorithm,a surface reconstruction algorithm is proposed to reconstruct the garment model which takes an undirected point cloud as input and generates interpolated surface by triangulation.The method in this paper is compared with the greedy projection triangulation method and the Poisson method.The results show that the garment model reconstructed by our method shows better performance.The obtained meshes are evenly distributed,the patches are smooth,and the original shape of the garment point cloud is retained.