Research On Feature Point Automatic Extraction And Model Reconstruction Of Auto Front-end

As a new product development technique, reverse engineering can improves modelquality and shorts product developing period. Reverse engineering is playing anincreasingly important role in auto industry and provides a new efficient method for modelreconstruction of automobile panel with complicated shape. The automatically extractingfeature point of scattered data and the reconstruction of auto front-end panel model andhole repairing are the main research components of this paper.Data segmentation is an important guaranty for model reconstruction. In most case,the characteristics of scattered data are not obvious and points distribute irregularly, whichbrings great difficulties to data segmentation. In this article, a method is proposed forautomatic feature point extraction coping with scattered data to this question. The stepsand specific algorithm of feature point extraction are given. With Visual C++and OpenGL,a system is developed for extracting feature point. Graph transformation and feature pointsextraction are two main modules of the system, which is proved to be useful for efficiencyand quality of data segmentation by two examples of feature point extraction.The model of auto front-end is obtained from scattered points with reverse softwareand a generic way is provided for reconstructing model with complex compound surface.This paper includes some common issues in reconstructing model and offers methods toresolve these problems, as non-quadrilateral hole repairing and pairwise rounding amongthree surfaces. The commonly surface trimming in Imageware is described. Finally, thehole repairing theory is proposed and three methods for repairing data are analyzed in thisarticle. The surface hole of front-end panel model is repaired, which establishes thefoundation for subsequent improvement and design.