用神经网络法实现反向工程中的曲面插补

提出了一种用神经网络方法对反向工程中曲面的坐标．点进行插补，将插补点与NURBS计算结果进行了比较，结果证明这是一和非常有效的构造曲面方法．并对这种方法进行了探讨。

A neural network approach is developed for surface reconstruction in reverse engineering. The calculating results of the neural network are discusses and compared with those of NURBS algorithm.This method is proved to be efffective for surface construction.