针对密集点云的快速自适应四边形网格生成算法

为了能够从密集点云直接获得四边形网格,而不需要通过三角形网格重构获得,提出针对密集点直接构造的四边形网格生成算法.首先进行点云数据体素化得到体素模型,建立体素和点云的索引关系,并对体素做精细化操作,以提高映射效果;然后通过体素模型外表面的顶点与原始点云的映射得到四边形网格模型,并对四边形网格进行优化.在斯坦福的数据集上进行实验,并使用MeshLab软件进行效果展示,结果表明,该算法可以基于密集点云直接生成四边形网格模型,同时可以通过调整体素大小来自适应地改变算法效率和四边形网格的大小.

Rapid and Adaptive Quadrilateral Mesh Generation Algorithm from Dense Point Cloud

For that quadrilateral meshes can be directly generate from dense point cloud instead of remeshing based on triangular meshes,this paper proposes a novel method which is able to achieve quadrilateral meshes straightly from dense point cloud.Firstly,this paper implements the point cloud voxelization so that the indices between voxels and points are established,and refines voxels to improve the performance of mapping.Then,the quadrilateral mesh model is attained by mapping between vertices of the external voxel surface and regularizing the quad mesh.At last,Stanford datasets are tested and showed by MeshLab,and the results demonstrate that the algorithm can immediately generate quadrilateral mesh model based on the dense point cloud,and adaptively change the efficiency of algorithm and the size quad meshes by adjusting the size of voxels.