Research On High-percision Reconstruction Of G~1 Continuous Cross Section Data In Reverse Engineering

Two high-precision method oriented feature are proposed to solve the non-ideal cross section feature reconstruction, because the boundary constraint is imprecise, which is caused by the inaccurate segment point in reverse engineering. The obtained section data is preprocessed, and then reconstructed accurately. Because of high accuracy recognition of the segment points, it makes the reconstruction results can effectively restore the product model of initial design intention, thus it improves the reconstruction accuracy and quality of three-dimensional CAD model.The reconstruction issues of line and circular and free-form curve are discussed. The line and circular arc are represented by algebraic calculations, and the free-form curve is represented by parametric B-spline curve. Their least-squares fitting models of curves and the representations of geometric constraints between curve features are given out. The existing issue of reconstruction technology of cross section data is analyzed.A high accuracy reconstruction method of cross section data is presented based on one-dimensional search. According to the discrete curvature of the cross section data information, extract segment points preliminary, determine the ideal region of segment points, and divide sectional data into single feature data segment by a regional endpoint. Reconstruct the line first, and then reconstruct free features. Construct optimization model based on boundary constraint model and free feature reconstruction model, use the Lagrange multiplier method to solve this optimization model. In the process of free feature reconstruction, extraction model of precise connection points between features is set up, using the golden section method dynamic to search the optimal point.A high accuracy reconstruction method of cross section data is presented based on one-dimensional search. We first determine the area of the ideal segment point based on the curvature information of data points. Then we use common gird method to search for the optimal segment point dynamically. During the search, for every grid node, the endpoint of line segment interpolates it, and the corresponding endpoint of free-form feature also interpolates it while satisfying G1 continuity. Meanwhile, both the sum of approximation error from all points to fitted curve and the number of control points of B-spline curve are counted to select the optimal segment point. Finally, reconstructing the cross section feature satisfied the boundary constraints based on the optimal segment point.The proposed high-precision reconstruction methods of G1 continuous cross section data based one-dimensional and two-dimensional searches have been applied in the reverse engineering CAD software, STLViewer. Through a re-modeling example in STLViewer, the feasibility and applicability of the proposed methods in the industrial production CAD re-modeling are demonstrated.In the end, we conclude the high-precision reconstruction methods of G1 continuous cross section data, and some advance topics for future work are addressed and discussed briefly.