Title :
LS-RBF network based 3D surface reconstruction method
Author :
Wen, P.Z. ; Wu, X.J. ; Zhu, Y. ; Peng, X.W.
Author_Institution :
Guilin Univ. of Electron. Technol., Guilin, China
Abstract :
We propose a new method for surface reconstruction from scattered point set based on least square radial basis function network in this paper. The RBF network is trained by fewer samples and we can get the weights of this network. Then an implicit continuous function is constructed to represent a 3D model. In this method, a binary tree is used to efficiently traversal the data set. Our scheme can overcome the numerical ill-conditioning of coefficient matrix and over-fitting problem. Some examples are presented to show the effectiveness of out algorithm in 2D and 3D cases. The numerical experiment shows high efficiency and satisfactory visual quality.
Keywords :
computer graphics; matrix algebra; radial basis function networks; surface reconstruction; trees (mathematics); 3D model; 3D surface reconstruction method; LS-RBF network; binary tree; coefficient matrix; computer graphics; implicit continuous function; least square radial basis function; neural network; numerical ill-conditioning; overfitting problem; visual quality; Binary trees; Electronic mail; Filtering; Interpolation; Least squares methods; Neural networks; Radial basis function networks; Reconstruction algorithms; Scattering; Surface reconstruction; Neural network; Radial basis function; point filtering; surface reconstruction;
Conference_Titel :
Control and Decision Conference, 2009. CCDC '09. Chinese
Conference_Location :
Guilin
Print_ISBN :
978-1-4244-2722-2
Electronic_ISBN :
978-1-4244-2723-9
DOI :
10.1109/CCDC.2009.5195232
