Neural Network To Surface Reconstruction Based On CAA Optimization Technology

With the development of CAD/CAE/CAM technology, the application of reverse engineering is constantly expanding, the surface reconstruction technology is the key in reverse engineering, which has been more and more attention. In order to get high precision of the original surface, some scholars have designed a lot of different algorithms, although there is a part of the mature algorithm,but defects is still existed. Aiming at the deficiency of this part, a new surface reconstruction based on neural network technology is appearing now in this paper.Based on BP and RBF neural network theoretical research deeply, this paper research maked many modeling and simulation experiments on noise point elimination and data recovery of point cloud, combined with CATIA secondary development technology, using CAA platform realized the combination of neural network algorithm and CATIA software effectively. It realized the application of noise point elimination algorithm in the CATIA software without the Matlab environment, making up deficiencies of point cloud pretreatment in the CATIA software and cutting down the time of reverse engineering, improving the the precision of surface reconstruction, eventually achieving the goal of surface optimization.