Implicit Surface Reconstruction from Scattered Point Data with Noise

Abstract

Abstract. This paper addresses the problem of reconstructing implicit function from point clouds with noise and outliers acquired with 3D scanners. We introduce a filtering operator based on mean shift scheme, which shift each point to local maximum of kernel density function, resulting in suppression of noise with different amplitudes and removal of outliers. The "clean" data points are then divided into subdomains using an adaptive octree subdivision method, and a local radial basis function is constructed at each octree leaf cell. Finally, we blend these local shape functions together with partition of unity to approximate the entire global domain. Numerical experiments demonstrate robust and high quality performance of the proposed method in processing a great variety of 3D reconstruction from point clouds containing noise and outliers.

An iterative slicing reconstruction method for point cloud surface holes

“…For the hole repair problem of scattered point clouds, the algorithms can be divided into the following three main types: repair methods based on discrete voxels, methods based on triangular dissection, and repair methods for constructing implicit surfaces.Firstly, the voxel-based repair method [4][5] is usually used to repair holes with small local area, and it is very easy to distort. Secondly, the triangulation-based repair method [6][7][8][9][10] usually performs initial triangulation on the point cloud surface holes directly, and then re-optimizes the mesh obtained from the initial triangulation to achieve the repair results, which is a simple algorithm, but the repair results of this method are often not satisfactory when the point cloud data are complex surfaces. Finally, the implicit surface fitting-based repair method [10] is the most common method in the point cloud hole repair algorithm, which extracts the point cloud hole boundary information to fit the repair surface to cover the holes, and then achieves the repair effect by resampling on the surface, but for point clouds containing multiple surfaces or uneven density, this overall surface sheet fitting repair method has certain defects.In recent years, many scholars have conducted a lot of researches, such as Yang et al [12] used least squares support vector machines to build surfaces to achieve point cloud hole repair, which can effectively repair point clouds with smooth surfaces, but cannot repair point clouds with large curvature changes due to the lack of feature information; He Dongjian et al [13] proposed to use three times B spline curve for hole repair, which has a good effect on repairing large area holes, but repairing Gai et al [14] proposed the repair of point cloud holes based on the motion structure, and the method has good effect on repairing complex texture of point cloud holes, but poor effect on the other hand; Hararyg et al [15] proposed the method of repairing holes by matching the original point cloud and hole features, and using the feature information of the point cloud to improve the matching degree to improve the hole repair accuracy.…”