An Adaptive Sampling Approach for Digitizing Unknown Free-form Surfaces Based on Advanced Path Detecting
Keqing Lu,Wen Wang,Yaofeng Wu,Yanding Wei,Zichen Chen
DOI: https://doi.org/10.1016/j.procir.2013.08.034
2010-01-01
Abstract:Reverse engineering technology has been widely recognized as a crucial strategy for innovative design of products. However, currently one of major problems for its application is that huge amounts of digitized points cloud data need to be processed for geometrical surface modelling, especially for unknown free-form surfaces which are widely used in industry fields such as aircraft, automobile, shipbuilding, steam turbine, etc. In order to digitize unknown free-form surfaces accurately and efficiently, an adaptive sampling approach based on advanced path detecting has been presented. The method is characterized by integrating one touch- trigger probe and two laser probes onto the same Z axis of Coordinate Measuring Machine (CMM), and the two laser probes are located on both sides of the touch probe. During sampling, the CMM scans the free-form surface line by line. At the scanning direction, the laser probes are applied to collect the measuring path information in front of the touch probe. And the data acquired by the laser probes is processed and fitted into B-spline curves real-timely, and then the curvature feature of the measuring path is extracted. The obtained path information is used to generate a new measuring path for touch probe to achieve adaptive sampling. While at the cross-section feeding direction, curvature analysis algorithm based on the angle variation between measuring sections is employed to subdivide the cross-sections when needed. Repeating this procedure, the measuring sections will be distributed adaptively. Simulation and experiment results indicate that the sampling points generated by the touch probe are laid out rationally according to the curvature of the surfaces, therefore the sampling efficiency and accuracy can be improved significantly by the proposed approach.