Research On Fast Adaptive Slicing Algorithm Of 3D Printing Based On Point Cloud

3D printing is a kind of technology which is used to form model by additive manufacturing. One of the general processes of 3D printing is as follows: First, the model data is acquired through a 3D laser scanner, then the model is divided into layer information through a fixed hierarchical algorithm, finally the model is printed step by step according to the layer information. With the development of 3D digital laser scanning and 3D printing technology, the size of the cloud data which needs to be processed is increasing, the accuracy of the model which needs to be dealt with is also becoming higher and higher, and there are a lot of noise in the data. Hence, the efficiency of data processing cannot be ignored.Nowadays most hierarchical algorithms generate triangle facets information based on the point cloud data, and then produce the slice information. In this thesis, a fast self-adaptive thickness adjustment slicing algorithm based on the point cloud data is presented, which could avoid the surface reconstruction(modeling) and the slicing error can be minimum when the slice layer is relatively small. The slicing algorithm skips the modeling process, adjusts the layer thickness based on the point cloud data directly, and then extracts the contour, which significantly improves the 3D printing efficiency.First, the present research status of layering thickness adjustment algorithm is analyzed, and a method which adjusts the thickness constantly and calculates the model error to meet the best layer thickness is proposed in this thesis. By analyzing the cause of the error, the relationship between the layer thickness and the model error caused by the staircase effect is inferred. Using the data structure of “Pre-grouping plus quad-tree” to store the data and quick sort to pre-group processing, which enhances the efficiency to search a layer of point cloud data at any interval. A layer of point cloud data is stored based on quad-tree basic data structure, which is conducive to analyze the influence of changes in the quad-tree leaf node on model errors. Experimental results show that the algorithm is high in time efficiency, small in the error, and has a good effect on the complex geometric features data model.Second, a method based on self-adaptive thickness adjustment slicing algorithm is proposed to extract contour in this thesis. In order to speed up the retrieval and traversal for point cloud data, quad-tree data structure is used to store the data. a combined method based on projection and quad-tree leaf node to extract slice data is used to reduce the amount of data to be processed significantly. The characteristics of quad-tree itself lead to there is some redundancy in the new generating data which could cause some errors for polygon feature extraction, therefore, a further repair based on neighboring relationship is needed to improve the effectiveness of the slice data. The adjacent point sequence is established by the slice data, then the characteristic polygon is determined, and the cross phenomenon existed in the characteristic polygon is repaired. Finally the three B-spline algorithm is used to fitting the curve.