Based on Scattered Data Surface Reconstruction

With modern design and manufacturing techniques and computer technology, reverse engineering in the field of mechanical engineering gained more and more widely . Three-dimensional surface reconstruction from scattered data points in reverse engineering technology is a very important research topic . In this paper, three-dimensional surface reconstruction from scattered data the actual needs of the scattered data preprocessing and surface reconstruction techniques were studied and discussed. On the one hand for mass scattered points , we propose a simplified algorithm for non-uniform , the other is based on Delaunay Triangulation Algorithm -Cocone surface reconstruction algorithm proposed improvements. Scattered non-uniform three-dimensional simplification of the data , the first data set to the minimum size of the bounding box is divided into the same number of cubes , and at each data point of the cube containing it contains the surface calculated within the local approach to change , and according to a local law set the threshold to change the data set is non-uniform simplified. Based on surface reconstruction of scattered data , in the study of the Voronoi diagram and Delaunay many algorithms , select Cocone algorithm as our theoretical basis for Cocone algorithm boundary treatment deficiencies, it presents itself as a local process improvement methods: The distance conditions triangle filtering. This method can quickly get a grid that can better reflect the topology of the surface to be built and geometry information . Experimental results show that the improved method is very effective in the detection boundary , the boundary can be a good filter to those on the unconstrained Delaunay triangles .