Research On Reverse Modeling Of Point Cloud Data Based On Feature Analysis

With the rapid development of computer vision related technology,3D scanning and reverse engineering have become a new sunrise industry.Reverse modeling is no longer a unique technology in the traditional industrial field,and it is gradually moving towards a life-oriented and service-oriented application scene.However,the current reverse modeling technology is mainly based on drawings and 3D modeling software,mainly manual,inefficient and time-consuming.Moreover,there is a certain technical threshold in the operation of 3D modeling software,and there is a large shortage of professionals,which limits the efficient application of reverse modeling technology in engineering.In order to solve the above problems,a point cloud data reverse modeling algorithm based on feature analysis is proposed.The algorithm is simple and efficient,which can reduce the number of triangular patches and retain local features,so as to ensure the quality of the reconstruction model and effectively solve the problem of reverse modeling of point cloud data.The main work of this paper is as follows.(1)Reverse modeling based on the scanned point cloud data,through voxel filtering,statistical filtering and other methods to simplify point cloud data,eliminate outliers,reduce the amount of point cloud data,in order to improve the speed of subsequent calculation;for details such as curved surface bends,the interpolation algorithm is used to fill the local sparse point cloud to improve the quality of the point cloud model and improve the reconstruction effect of detail features.(2)Through the feature analysis of massive point clouds,the feature points which can represent the key geometric information are extracted and labeled.Retain as much as possible the original point cloud in the neighborhood of the feature points,so as to retain the detailed information and improve the local quality of the reconstruction model,especially for the reconstructed objects with complex structure,and simplify the point cloud again in the area where the feature is not obvious.In order to reduce the volume of the model and improve the loading speed in the actual project.(3)The region growth algorithm is used to segment the point cloud,and several sub-regions with similar feature measures are obtained.Select the appropriate reconstruction parameters for the point cloud data of different regions,and combine the triangulation algorithm for surface reconstruction to further improve the quality and accuracy of reverse modeling.(4)In the neighborhood of feature points,the mesh is optimized by bridging and filling based on dense point cloud data.The reverse modeling experiment based on feature analysis is carried out on the point cloud data of a differential pressure transmitter,and the generated 3D model is compared with the 3D modeling software Geomagic and the reverse modeling results based on curvature features only.It is verified that the surface of the 3D model reconstructed by this algorithm is continuous,closed,the quality is better,the number of triangular patches is less,the volume of the model is small,the operation efficiency is high,and the time-consuming is less.The effectiveness of the algorithm is verified.