The Processing And Application Research Of 3D Laser Point Cloud Data

3D point cloud data can provide a real 3D city model for 3D geographic information system,simulation and virtual technology,and the construction of digital city.It also provides simulation and Simulation for environmental design,planning and design,traffic command,signal station,environmental analysis,virtual tourism and other works related to the city.There have been many years of research history and a lot of research achievements made by the 3D model of point cloud reconstruction and related processing technology.However,because of the different technologies for obtaining 3D point cloud data in various fields,the existing point cloud processing technology has shortcomings in the application of various fields,such as noise,external points,data loss,loss of sharp features and uneven sampling,etc.The robustness is insufficient.Many research institutions and researchers have been engaged in the research and improvement of key technologies in point processing to improve the quality and processing speed of 3D point cloud data.The main purpose of this study is to improve the processing efficiency of point cloud data in the construction of urban 3D model,and to provide some support for the rapid construction of digital cities by improving the processing efficiency of point cloud data and the accuracy of model reconstruction.Based on the analysis of the theory and application of 3D reconstruction based on 3D laser scanning data,this paper summarizes the key technologies of the surface 3D laser scanner to build the model,and summarizes the applied theory and model reconstruction methods.On the basis of reverse engineering technology,using Trimble Real Works to preprocess the data of point cloud scanned by Trimble TX8 intelligent laser scanner,the following four aspects are carried out:(1)by considering point space of point cloud data,weighting analysis and combining weighted normal vector with WLOP algorithm,a new En WLOP method is used for point cloud data.Simplification;(2)smoother than jet smoothing and bilateral filtering,using the optimal bilateral filtering of the building to smooth the point cloud data and its normal vector,and get the high quality point cloud edge.(3)through the Poisson algorithm,the point cloud data is quickly rebuilt and saved time by the Poisson algorithm,and(4)finally through MFC combined with Open GL Methods the development of point cloud data processing software is targeted.The main results are as follows:(1)a WLOP algorithm combined with weighted normal vector is designed to improve the precision and speed of the point cloud simplification by comparing with the results of the mainstream software.(2)most of the point cloud processing software at the present stage does not consider the processing method of the normal vector,estimates the normal vector,improves the accuracy of the modeling,and(3)smoothing the 3 point cloud data.The method is compared,and the bilateral filtering smoothing method is the best way to smooth the building.(4)because the processing of the point cloud data is scattered in the mainstream software,there is no special software for the building,this paper designs and develops the point cloud location software for the building,which provides support for the construction of digital city.