Research And Application On Parameter Calculation Of Complex 3D Model Based On Point Cloud

In the manufacturing industry,there is an inevitable deviation between the final product and the theoretical design model due to manufacturing errors.In order to ensure that the product meets the design requirements,the quality inspection of the final product is essential,especially those important parts,and the inspection contents can usually be attributed to the measurement of certain geometric and physical parameters.Conventional entity-based measurement has the disadvantages of equipment dependence,low efficiency and low precision.With the development of 3D measurement technology,Point cloud can reflect the shapes of real models precisely,Besides,it has the advantage of fast scanning with a large amount of information.Therefore,3D scanning measurement has been widely used in the inspection of complex model,but the extraction of 3D parameters based on point cloud is still a problem to be further studied.In this study,the point cloud obtained by 3D scanning is taken as the research object,and a parameter calculation scheme based on point cloud is proposed,which can be used to calculate the geometric and physical parameters of the real model.In the scheme,the 3D parameters calculation is transformed into a 2D problem by slicing,and the reconstruction of the total model is equivalent to a series of curve reconstruction.In the slicing process,the virtual edge approach is used to obtain the refined cross-section point set,and a simplification algorithm based on RNearest neighbor is proposed to homogenize the point set.In the curve reconstruction step,the algorithm based on computational geometry is introduced to generates cross-section contour.For the reconstructed section contours,an improved approach based on triangulation segmentation is proposed,which can automatically calculate the section parameters accurately by identifying the inner and outer contours.In order to optimize the efficiency of massive point cloud processing,an adaptive slicing algorithm based on cross-sectional variation is proposed.The evaluation criteria combining area change and centroid offset are used to identify the point cloud model in a quantitative way.After experimental comparison,under the same calculation accuracy,the adaptive slicing algorithm can reduce the times of slicing and improve the efficiency of calculation.Based on the secondary development platform NX,the computational geometry library CGAL,and the point cloud library PCL,a 3D geometric and physical parameters extraction software based on point cloud is developed by implementing the above parameter calculation algorithm.Taking the dynamic balance detection of automobile crankshaft and the cross-sectional area detection of aviation engine as the application examples,the feasibility and practicability of the scheme are verified.