Data Reduction Based on Bi-Directional Point Cloud Slicing for Reverse Engineering

Abstract

In reverse engineering, complex free-form shaped parts are usually digitized quickly and accurately using the newly arisen non-contact measuring methods. However, they produce extremely dense point data at great rate. Not all the point data are necessary for generating a surface CAD model. Moreover, owing to inefficiencies in storing and manipulating them it takes a long time to generate a surface CAD model from the measured data. Therefore, an important task is to reduce the large amount of data. After analyzing the existing methods developed by other researchers, a new data reduction method, which based on bi-directional point cloud slicing, is presented in this paper. Using the proposed method, point cloud can be reduced while considering geometric features in both two parametric directions. Finally, a face model is used to verify the effectiveness of the proposed method and experimental results are given.