Research And Implement On The Algorithm Of Point Cloud Simplification And Fast Reconstruction

With the rapid development of three-dimensional scanning technology, the scattered points of the model can be accessed easily, which greatly facilitated the development of reverse engineering. Reverse engineering has now become a relatively independent research branch of the CAD / CAM system, and it is also one of the most active research points in the field of advanced manufacturing. And scattered points cloud simplification and triangular mesh reconstruction has an important foundation in the reverse engineering.This thesis focuses on the scattered points cloud simplification and triangular mesh reconstruction, the main research and achievements as follows:In the part of scattered point cloud data simplification, base on the research of existing algorithms, a local adaptive simplification method with feather reservation is presented for three-dimensional scattered points. This algorithm inherits the merits --- "feather reservation" according to keeping the feather points, and it can also avoid the complex calculation of curvature. Besides that, this algorithm can simplify the general points effectively, and does not appear "empty" phenomenon.In the part of triangular mesh fast reconstruction, the author compares the common method in the areas of triangular mesh surface reconstruction, to the shortcomings of having low efficient and incapable; a fast surface reconstruction algorithm based on muti-strategies is presented. The method combines the advantages of Projection-based reconstruction algorithm and Region Growing Triangulation algorithm. The result show that the triangular mesh has a fine form and it can keep the geometric feather and identify the holes and boundaries.Besides, a prototypical system based on these algorithms is implemented in this thesis, and several typical example are given as well. The result of these examples indicates the correctness and effectiveness of the algorithm above.