Research On Key Technology Of Building Surface Reconstruction With High Resolution 3D Point Cloud

The surface reconstruction method of the point cloud of building is a process of extracting the geometric structure of the object surface from 3D point cloud data and restoring the three-dimensional solid model.The algorithm for reconstructing the point cloud surface of building has currently broad application prospects in fields such as high precision urban mapping and virtual reality.Due to the diverse geometry of buildings,there are widespread problems in traditional reconstruction algorithms,e.g.,low fitting precision,and incompleteness of building structures.The proposed algorithm focuses on exploring the effective algorithm of vectorized surfaces of building surface to improve the accuracy of the reconstruction and enhance the closedness of the model.1.In view of the problem of fitting based on 3D point cloud data with noise or missing,a algorithm based on weighted constraints for reconstructing the point cloud surface of building is proposed.Firstly,the adaptive weights are designed according to different types of point cloud data in the data surface initialization process.Secondly,and the optimal weight function in the model fitting is selected to get the smallest data-fit error.Finally,a algorithm of surface reconstruction based on the regular set is constructed to achieved extraction based on weighted constraint of building surface.Experimental results of reconstruction of buildings with different data sources and different scales indicate that the proposed algorithm can obtain higher precision of building surface model on the point cloud data with high noise and low precision.2.Focusing on closure of reconstruction model,a algorithm based on closed constraint of polygon for reconstructing the point cloud surface is proposed.Firstly,point cloud data is initially fitted by the method of Principal Component Analysis.Secondly,candidate set is generated by expanding the initial primitive set as the method of the regular set reconstruction.Finally,a closed constraint is added to the energy minimization equation containing the data item,the regular term,and the smoothed term to select the final model plane set.Compared with the algorithm based on regular set for reconstructing the point cloud surface of building,the algorithm based on closed constraint of polygon for reconstructing the point cloud surface is robust to noise and can effectively reconstruct the lightweight model with closed boundary.In this paper,considering low fitting accuracy and poor structural closure of reconstruction models,a algorithm based on weighted constraints and closed constraints for reconstructing the point cloud surface of building is proposed.By analyzing the influence of fitting weights ofmulti-class data and closure of model boundary,the experimental results of building reconstruction with noise,different scales and different structural complexity show that the proposed algorithm effectively improves the model fitting accuracy and closure,and obtains a high-precision and compact building model.