Research On Processing Technology Of Point Cloud Model With Surface Features In Reverse Engineering

With the gradual completion of reverse engineering technology,its application fields are becoming more and more extensive.Combined with the rapid development of 3D image processing technology,the technology related to point cloud data processing has also been further improved.For some data models with free-form surfaces,the model information can be obtained by collecting point cloud data,thus reducing the cost of product design,development and subsequent optimization.However,the current processing methods for point cloud models have not been studied for the processing of point clouds with surface-like features.At present,there are few boundary features identified in the surface point cloud model,distortion in the refined boundary of the residual model,reconstructing the model and forming verification,and providing data guarantee for the subsequent processing technology.In this topic,a series of researches are conducted based on the construction of neighborhood spatial relations for point cloud data.There are two major parts of the research: the boundary and neighborhood feature identification and data refinement research for thin-walled class of defective point cloud data with surface features and the reconstructing strategy and molding experimental research for closed surface solid point cloud models,as follows:(1)A boundary feature detection and its neighborhood data recognition algorithm is proposed for the thin-walled class surface point cloud model,which uses the normal vector angle of the point cloud as the feature detection operator,and then uses the Euclidean distance to search the feature nearest neighbor data points to achieve the recognition of the disabled model boundary and neighborhood feature points.The method can identify and retain more boundary feature data points based on the recognition of boundary feature points based on normal vectors to reduce the recognition error and improve the efficiency and accuracy of subsequent processing.(2)At present,most point cloud data reduction methods are reduction methods for complete models.However,if the data is lost,a hole phenomenon occurs.This requires that when streamlining point cloud data,not only the original boundary feature points of the model,but also the feature points in the neighborhood of holes should be preserved,so as to facilitate subsequent point cloud hole repairing,etc.To this end,a data reduction method is proposed for the thin-walled incomplete surface point cloud model to retain its boundary and feature points in the neighborhood of holes.The average curvature is used as the data reduction threshold to simplify the data of non-feature areas.Finally,the extracted feature data points are fused with the simplified non-feature data points to achieve the maximum possible streamlining of non-feature points and ensure the boundaries and their holes.The neighborhood features are not lost,and the reduced standard deviation and surface area change rate are better than those of random reduction and curvature reduction.(3)Aiming at the research on the reconstruction strategy and molding test of closed surface point cloud model,taking massager as an example,with the help of 3D printing technology,the rapid design and manufacture of plastic products is realized,and the reconstruction strategy and molding test for plastic products with complex curved surfaces are summarized.technical points in.First,use a laser scanner and Geomagic Wrap to complete the collection of entity point cloud data and the preprocessing of model data.Secondly,with the help of Geomagic Design X software,the reconstruction of the massager model is realized by the method of three-dimensional surface fitting.The surface entity is created by the method of surface fitting,which reduces the Boolean operation in the traditional model reconstruction method and reconstructs its original surface geometric features to the greatest extent.The error comparison between the reconstructed solid model and the collected and processed point cloud data verifies the feasibility of the model reconstruction method.Finally,through the analysis of the currently commonly used 3D printing process,the 3D printer with fused deposition process is selected,combined with the application of the model in practice,the product trial production is carried out from three different molding directions.