In several applications, the only available information on a surface is a set of points lying on it. Most computations on this surface then require to find first a piecewise linear approximation of the surface from the sample points: this is the problem of surface reconstruction.
This web site presents a new algorithm developed by David Cohen-Steiner and Frank Da from the Prisme Research Group at INRIA. Experiments show that the algorithm is very fast, and achieves topologically correct reconstruction in most cases. Moreover, it can handle surfaces with complex topology, boundaries, and non uniform sampling.
You are invited to test the algorithm, as it is available as a Reconstruction Service.