Surface reconstruction from scattered point via RBF interpolation on GPU

In this paper we describe a parallel implicit method based on radial basis functions (RBF) for surface reconstruction. Practical applicability of RBF methods is hindered by their high computational demand, that requires the solution of linear systems of size equal to the number of data points. The implementation of our method relies on parallel scientific libraries and is designed for exploiting Graphic Processor Units (GPUs) acceleration. The performance of the proposed method in terms of accuracy of the reconstruction and computing time shows that RBF interpolation can be very effective for large scale surface reconstruction problems.