法向估计的屏蔽泊松算法三维点云重建

Three-dimensional point cloud reconstruction of normal estimated screened poisson algorithm

针对屏蔽泊松算法在表面边缘易生成错误曲面的不足,提出一种基于法向估计的屏蔽泊松算法。该算法首先对扫描的点云数据进行位置和梯度约束;然后,利用法向估计来获取朝向一致的法向量,进而得到点云的向量场;最终,求解屏蔽泊松方程,通过提取点云的等值面从而来重建模型表面。在重建的三维模型上评测新算法,并与其他几种重建方法进行比较。结果证明,所提算法构建的三维模型整体上提高了法向量精确度,能有效减少非封闭边缘的伪曲面重建,同时对带孔洞的表面点云有良好地填充能力,对表面信息丰富的点云模型,能使纹理细节更显著。

Focus on the issue that Poisson equation in 3D point cloud reconstruction,a screened Poisson based on normal estimation is proposed.Firstly,the position and gradient constraints are performed on the scanned point cloud data.And then,method to estimate the point cloud to obtain consistent towards normal vector and points cloud vector field is obtained;Finally,the screened Poisson equation is solved,and the model surface is reconstructed by extracting the iso-surface of the point cloud.In this paper,the method of 3D model reconstruction is evaluated and compared with other reconstruction methods.The experimental results show that the 3D model constructed in this paper can improve the accuracy of the normal vector as a whole,can effectively reduce the pseudo-surface of not-water-tight edge,has a excellent hole repairing mesh,and the higher-quality surface reconstruction,point cloud data can make the texture detail more remarkable.