利用RBF神经网络实现三角网格模型的孔洞修补

Using RBF Neural Network to Realize the Hole Repair of Triangular Mesh Model

本文提出一种利用径向基函数(RBF)神经网络实现三角网格曲面孔洞的修补算法,首先用孔洞边界周围的三角片顶点作为学习样本训练RBF网络,然后对孔洞进行平面填充,获得新增三角片的顶点,最后用已训练好的RBF网络将其优化,实现三角网格孔洞的修补。

This paper presents a algorithm for the reconstruction of triangular mesh surfaces using the radial basis function （RBF） neural network. Firstly, the RBF network is trained by using the triangle mesh vertices around the hole boundary as studying samples. Then the holes are filled in the hole, and the added vertices are obtained. Finally, thetrained RBF network is used to optimize the mesh, and realize the repair of triangular mesh.