Surface Reconstruction from Unorganized Points

1. Briefly summarize the paper’s contributions. Does it address a new problem? Does it present a new approach? Does it show new types of results?  [AS] This paper presents an algorithm that produces a simplicial surface, approximating a manifold surface M, from an unorganized set of points containing little to no information about M. It presents a new approach to the problem of recovering surfaces from range scans and other practical situations.  [DS] The method takes as input an unorganized set of points and outputs a simplicial surface that approximates a manifold. Makes no assumptions on the topology, presence of boundaries or the geometry of the manifold.  [FP] The authors propose a method to reconstruct a surface as the zero set of a signed distance function. They locally approximate the surface by fitting tangent planes at point neighborhoods and defining a coherent normal orientation. Their approach is valid for any raw set of points (no normal required), independent of the acquisition source. They present results in point sets obtained from meshes, ray tracing, range images and contours (horizontal slices).  [JD] The paper introduces a method for general surface reconstruction from unorganized, unoriented points. The algorithm assumes nothing about the data and reconstructs surfaces of arbitrary topology with possible boundaries. It is the first use of moving least squares for surface reconstruction.  [LF] The paper presents a method that can reconstruct surfaces of arbitrary topology unlike parametric methods, only requires point cloud data (no surface normals needed), and handles boundaries more naturally than implicit methods as specified by the authors.  [MK] This paper proposes one of the first methods for implicit surface reconstruction from point-sets. The paper makes several contributions: (1) it describes an approach for consistently estimating normals for unoriented points and (2) it shows how to use the oriented points to construct a piecewise linear (though discontinuous along boundaries) approximation to the distance function.