Preprint B15/2007
Reconstrução de Regiões a partir de Amostras com Ruído
Emilio Vital Brazil
Keywords: Noisy Samples; Function Approximation; Radial Basis Functions
Information extraction from point clouds is a classic problem in fields as diverse as statistics, computer vision, geometric modeling, reverse engineering and computational geometry. In this work, we are interested in point clouds representing a bounded region of the plane defined by a finite set of simple closed curves. We approach this problem as one of function approximation. We designed a reconstruction algorithm based on radial basis functions that presented interesting results in conformance with the theory developed for its construction.