Study On Key Techniques In Reverse Engineering CAD Modeling Of Complex Parts

In this dissertation, key techniques in reverse engineering CAD modeling of complex parts are systematically studied.The article introduces fundamental conception of reverse engineering, application areas of reverse engineering and steps of CAD modeling. Key techniques in reverse engineering CAD modeling and its development status are also introduced. At the same time , introduce the feature applications in geometric modeling and the significance of feature introducing in reverse engineering CAD modeling.In allusion to the feature of reverse engineering and the function of geometric feature and constraint in reverse engineering. The definition of feature and constraint reverse engineering CAD modeling is systematically studied. The features are distributed into four kinds including point, line, surface, voxel and two feature spaces including main feature and auxiliary feature. In allusion to mechanical parts, three kinds of geometric constraint including line-line constraint, surface-surface constraint and voxel-voxel constraint are introduced.With regard to the analysis of the cloud data in reverse engineering, an effective method for unorganized cloud data smoothing and completing is presented. On the basis of triangulation, set up local coordinate in each triangulation area and use a normal distribution model in the theory of probability to delete the noise points partly. Then make B-B surface on each triangulation net to fair data globally. In allusion to the phenomenon that partial data can't be measured in actual data surveys, a data-completing method is proposed. Estimate the geometry information of the acquired data points which are around the unknown data points. Then set up a optimized energy model on the basis of aforementioned geometry information to complete data.With regard to the extraction of quadric surface feature, data points' curvature and normal are calculated firstly. Regard them as the gist to extract several kinds of simple quadric surface feature. New algorithms for fitting plane, sphere, cylinder and cone with LSM are discussed.With regard to the extraction of free-form surface feature, a slicing algorithm, used to get section data from point data is presented. The method to extract feature points from section data is also studied. And these feature points are regarded as the gist to recognize section feature. When the section feature is recognized, they shouldbe fitted with LSM, and then skinning method is used to reconstruct free-form surface modeling.In the summary of this dissertation, a number of advanced topics in reverse engineering CAD modeling is addressed for the future research.