Research On Surface Model Reconstruction And State Information Extraction Method Of Track Structure Based On Point Cloud Data

At present,the state detection technology of track structure in China has become increasingly mature,but the automatic modeling technology and the intelligent extraction technology of state information based on point cloud data are developing slowly in the field of railway operation and maintenance.Point cloud data contains high-precision three-dimensional coordinate information,which can truly describe the overall structure and morphological characteristics of the object.Based on point cloud data,it can realize rapid model building and state information extraction,with the advantages of high precision,high efficiency and full digitalization.Based on Point Cloud Library(PCL)and Geomagic Studio software,this thesis studies the surface model reconstruction and state information extraction of track structure by selecting the point cloud data of in-service track structure collected by 3D laser scanner as the research object.The main research contents are as follows :(1)Based on PCL and Geomagic Studio,the file format conversion,the simple visualization,the rotation pendulum and the filtering downsampling of track structure point cloud data are realized.These operations lay the foundation for the subsequent research on recognition and segmentation,surface reconstruction and state information extraction of track structure point cloud.(2)Based on PCL,the Euclidean clustering algorithm and region growing algorithm are used to recognize and segment the point cloud of the track structure components.And the recognition and segmentation effects of the two algorithms are compared and analyzed.Although the Euclidean clustering algorithm is simple and easy to use,the region growing algorithm has a better recognition and segmentation effect for the track point cloud data with connected parts.(3)Based on PCL,the greedy projection triangulation algorithm and the Poisson reconstruction algorithm are respectively implemented to reconstruct the surface model of the whole track structure and its components.And the surface reconstruction effects of the two algorithms are compared and analyzed.The greedy projection triangulation algorithm can clearly express the local detail features of the surface model,which is convenient to intuitively analyze the local state of the track structure through the model.The Poisson reconstruction algorithm can reconstruct a surface model with good integrity and smoothness,which is suitable for analyzing the overall state of the track structure.(4)Based on point cloud data,a track structure state information extraction algorithm is proposed.The extraction of fastener state information is realized based on the spatial position relationship of track structure fastener point cloud.Based on the thin slice point cloud data of the track structure geometric shape and position measurement position,the CAD list query is used to realize the rough extraction of the geometric shape and position information.The point cloud of rail section is intercepted,and the B-spline curve fitting algorithm is used to extract the rail profile curve.The research results of this thesis are helpful to promote the in-depth application of point cloud detection data in the operation and maintenance of railway infrastructure,and provide a fast and efficient method for railway management departments to obtain track structure status information.At the same time,it provides reference for the operation and maintenance of engineering fields such as highways,electric power and pipelines that need real and detailed feedback status information.