Model Reconstruction Of Urban Structures Based On 3D Laser Scanning

With the rapid development of city digitalization and intelligentization,the demand for urban spatial information data is no longer satisfied with simple two-dimensional data,and more real,more complete and more accurate three-dimensional mapping results are urgently needed.The 3D real scene model with visualization and measurability has been widely used in many fields of urban development.As a part of the city,the urban structures also need to construct more real,more sophisticated real model to meet the needs of social development.The key problem of real scene model reconstruction is to solve the contradiction between too much data and low work efficiency.As a new surveying and mapping technology,the ground 3D laser scanning technology is also called "360-degree scene replication technology",which can realize fast,efficient and non-contact collection of massive point cloud data of objects,the problem of collecting massive coordinate data is solved effectively.The point cloud data can be measured indoors after the data processing,and then produce a variety of surveying and mapping results.Therefore,3D laser scanning technology is widely used in ancient building restoration,archaeological mapping,cultural relics protection,digital city and film and television special effects.But how to deal with the massive point cloud data quickly and efficiently,how to obtain the high-quality point cloud,how to use the point cloud data to improve the efficiency of digital model reconstruction,is still the hot and difficult point of current research.In this paper,the application of 3D laser scanner in the reconstruction of urban structure model is studied,taking the gate of Dongwu University and the gate of XianCan ancestral hall as examples.This paper discusses the working principle and classification of the ground 3D laser scanner,analyzes the error source of the point cloud data,and tests the accuracy of the ground 3D laser scanner through the experiment,which lays a foundation for the following data collection work.Aiming at the problem of point cloud data processing,the methods of point cloud splicing,de-noising and simplification are studied.Two kinds of four splicing methods are tested,and the characteristics and adaptability of different splicing methods are analyzed.Taking the point cloud of the gate of Dongwu University as an example,the paper studies the Statistical Outlier Removal filter method and Geomagic Studio software.Finally,the VoxelGrid filter is used to simplify the data of the point cloud,and the high-quality overall point cloud data is obtained.In this paper,two different model reconstruction methods,irregular triangular grid method and multi-source data fusion method,are studied for different urban structures.For the single structure and the small volume of the structure,we can use the ground 3D laser scanner to scan the whole picture of the urban structure,and take the gate of Dongwu University as an example,use the irregular triangular grid method to reconstruct the model.However,in the social production practice,many structures are high,and it is difficult to collect the complete point cloud data at the top by using the scanner.In this paper,the gate of XianCan ancestral hall in Shengze is taken as an example,combining the advantages of ground 3D laser scanning technology and UAV tilt photogrammetry technology,a model with high accuracy and good feature of object details is constructed.