Research On Surface Intelligent Algorithm Scattered Points Reconstruction

Grey model is a traditional outdoor building decorative arts in Guang Zhou. The complex technology and unique artistic style make Grey model become one of the world’s intangible cultural heritage. But the permanent of save and industrial production for Grey model have been big troubles to plastic artists. With the constant progress of reverse engineering technology, It make a possible to solve the problem. Reverse engineering of the system structure is mainly composed of point cloud data acquisition, data processing, curved surface reconstruction and rapid manufacture and other several parts. However, surface reconstruction is the most important technology in it. The main stream of modern modeling technology is the surface reconstruction of NURBS curve. Owning to the complexity of gray plastic art form, it is difficult to use a single description of NURBS surface Grey plastic work. If using more than one, it would be a large work to certain between the processing surface,and it is easy to produce cracks.Aiming at the shortness of the NURBS surface, this paper proposed a kind of more suitable method for the Grey model to reconstruct work, an intelligent method of subdivision surface, which is controlling the low resolution grids and define control grid subdivision rules. So that it shows out a smooth surface, simple rules, topology adaptable. The main research content is as follows:(1) As for the data characteristics of Grey model, the author of this paper’scattered point cloud’, and the Delaunay triangulation has been studied. Delaunay triangulation has been the most popular used in pointing cloud data grid algorithm.It can be divided into two categories:one is to point cloud regional decomposition, each region of the special annotation, and each area of the triangle, and the algorithm of divide-and-conquer strategy; a kind of orderly arrangement of the point cloud data and will eat every point mark on the corresponding coordinates of the adjacent two points as the two points of the triangle, and then insert a point, a triangle, to repeat in between each neighboring points. By studying the characteristics of these two kinds of algorithms, this paper synthesize the two algorithms, and a certain degree of improvement, formed a point cloud triangulation algorithm, which is more efficient. After that, it can be in any area of the combination of Delaunay triangulation point gathers. Using this improved method, it can be more effectively realized, due to Grey plastic complex art form and complex point cloud data border area.In that case, too much difficult problems can be grid.(2) A method for the gray plastic mesh surface characteristics. By analyzing the surface characteristics of a large number of gray plastic, works point cloud grid lines special classification. The method of using the lowest energy consumption optimization vertex coordinate point cloud grid, it can deal with the important points of the grid point cloud optimization.After processing, it finally optimize point cloud in the form of linear equations to calculate. This method can improve the fit surface point cloud data grid. And treatment between gray plastic works is to optimize mesh quality of some subsequent processing,which will play an important role. If the grid is in poor quality, the final reconstruction Quality out of gray plastic surface will be reduced accordingly. Surface reconstruction will be more accurate expression to the artistic plaster work, plaster work so that there is more space in the future practical application in industrial use, reduce labor costs and accelerate the production of water-based.(3) put forward a Grey plastic work point cloud data features of smart subdivision surface algorithmGrey model works have such a good feature, which aout the local craft fine degree is high and the local area process is extremely simple shape to know each other. According to the characteristics of this kind of point cloud data, this paper proposes a smart since the inductive Loop subdivision method, using this method can effectively avoid met repetitive segmentation, by computing the curvature of the surface at any point, for all the curvature of the point cloud curvature similarity comparison, it can greatly reduce the energy waste produced by excessive similar regional segmentation and computational load, and at the same time, the area of similar point cloud can be smooth processing, avoid arise surface caused by the excessive similar to mark. Point cloud data of the subdivision surface singularity continuously in the border area, Grid subdivision method based on adaptive Loop presented in this paper, first of all, preprocessing of triangular mesh model for the triangular mesh vertex optimization, get the fitting control grid. And then to modify control in the grid vertex fitting, control of the vertex revised grid topology shape optimization, repeated iteration, generating meet the requirements of Loop subdivision surface.