Abstract:
The construction of a weldment surface from scattered data and extraction of welding seam is an important problem in welding robot stereovision system. The various methods on 3D surface reconstruction are reviewed at first, and according to weldment surface property, an autonomous, consecutive function approximation approach is presented to assign parameter values to obtained 3D points with noise for weldment surface by a least-square-fitting algorithm. Real range data are used to show the algorithm's performance and experimental results demonstrate the effectiveness of the proposed algorithm.
Key words:
Surface reconstruction Weldment surface Range images Least square fitting
* This project is supported by National Natural Science Foundation of China (No.59635160).
Selected from Proceedings of 2000 the First International Conference on
Mechanical Engineering. revised manuscript March 10, 2001
|