Surface reconstruction is a key technology in geometric reverse engineering. In order to get the geometric model of object we need a large number of metrical points to construct the surface. According to the strong points of RBF network such as robust, rehabilitating ability and approximating ability to any nonlinear function in arbitrary precision, we presented a new method to reconstruct B-spline surface by using RBF. Simulation experiments were made based on the theoretical analysis. The result indicated that this model could not only efficiently approximate incomplete surface with noise, automatically delete and repair the input wrong points through self-learning, but has a rapid learning speed, which improves the reconstructing efficiency and precision of dilapidation incomplete surface. The surface obtained by this model has a good smooth character. © Springer-Verlag Berlin Heidelberg 2005.
CITATION STYLE
Liu, X., Huang, H., & Xu, W. (2005). Approximate B-spline surface based on RBF neural networks. In Lecture Notes in Computer Science (Vol. 3514, pp. 995–1002). Springer Verlag. https://doi.org/10.1007/11428831_124
Mendeley helps you to discover research relevant for your work.