Research And Implementation On Key Technologies Of Surface Reconstruction Based On CMM Measuring Data

As one of the hot spots of the advanced manufacturing field, reverse engineering has recently been a relatively independent research branch in CAD/CAM systems. Its correlate area includes geometric measuring, image processing, computer vision, geometric modeling and digitalization manufacturing etc.Surface reconstruction is one of the key technologies in reverse engineering. The key technology is studied in detail in this paper. The technologies of data processing, NURBS surface reconstruction and the triangulation algorithm feature are researched in order to supply the perfect ways and means for the technique absorption, digestion and innovation of weapons products and build up foundation of developing the practical reverse engineering software.Firstly the data pre-processing technologies are researched, which include coordinate transform, coordinate fusion, data filtering. Under this premize, the method of data sorting based on layering dimension is put forward. With the CMM scanning data points, the layer dimension is defined, data layering and data sorting are performed. The topology relation is established. On the basis of data topology, the algorithm of data reducing which combines distance threshold with veltor angle is researched. The technology of feature extraction based on layering dimension is also studied, and all kinds of discontinuous points (jumping point, spinode, buckle point, curvature maximum and so on) are extracted completely. The technogogy of data segmentation based on the feature line frame is researched.On the basis of the research of layering dimension based data processing, a procedure is proposed that reconstructs Non-Uniform Rational B-Spline (NURBS) surface with features of measured data in this paper. With non-uniform data points, a method of accumulating chord length is presented to create non-uniform knots vector and guarantees the precision of the surface. An optimal algorithm for B-spline evaluation is put forward. The symbolic manipulation is used and a B-spline basis function is expressed as piecewise polynomials through coefficient calculation. The layering dimension is set to determine the minimum initial array. The temporary polynomial is exploited to save the intermediate result, and evaluations of polynomials are performed in the large-scale evaluation. In such a way, the computing velocity is greatly enhanced and the space usage is obtained through the determination of minimum initial array and the duplicate iteration cover of the...