The Reconstruction Of Curve And Surface From Noisy Sampling

The reconstruction of curves and surfaces is one of the important problems in reverse engineering. Sometimes, the reconstruction of surface can be changed into surve reconstruction.such as, surfaces of revolution, helical surface, profile surface, and so on.This paper focuses on curve and surface reconstruction from nosiy samplings and proposed the algorithm to get the curve and surface. In the algorithm of reconstruction, the refined neighbourhood can be constructed so that the data point is moved to a new position on this approximated curve,so the point cloud is thinner than old point cloud.Then the point cloud can be ordered. Finally, a smooth curve can be reconstructed by using curve fitting method. If the samples are sparse,a smooth surface is produced by interpolation.While the samples are dense,a number of small trianglar meshes will be produced.On one hand,it will produce redundant information. On the other hand, the characteristics of the sharp corners and sharp edges can't be reflucted when the curvature has a big change.In light of this situation; the approach of mesh optimization is generated for the small trianglar meshes by topology knowledge.Some examples are given here to show its validity and effectivity in reconstructing curves or surfaces with different shapes.