A STUDY ON DATA PRE-PROCESSING IN REVERSE ENGINEERING

摘要

Reverse engineering (RE) is an emerging technology that promises to play a role in reducing product development time. Data pre-processing is one of the most important technology in RE and directly decides the effect of the following works. That is, the result of data pro-processing directly influences the precision and efficiency of the CAD model rebuilding. In this paper, some of the key technology based on the scanning-line data would be detailed: (1) Noisy points canceling. Obvious noisy points of data points cloud would be cancelled by man-machine interaction using visual technology; (2) Data points zero division. Data points would be divided into different zero by curvature analysis on point clouds using MATLAB software programming. (3) Data reduction. The adaptive minimum distance method (AMD) is proposed to implement the reduction of the mass of scanning data points, selecting different minimum distance in different zero to reduce data points so that less data points are preserved at the less curvature zero and more points are preserved at the bigger curvature zero or sharp edge zero. (4) Data smoothing. The data points after reduction would be smoothed through a new kind of mixed filtering algorithm, composed of median filtering and weighting median filtering. And the algorithm principle and procedure of the whole data pre-processing is verified through practical application and some of experimental results are presented in the paper.