Research On Adaptive Reduction Point Cloud Block Modeling And Error Entropy Evaluation

With the rapid development of China's economy and the continuous acceleration of the process of social informatization,urban development has continued to shift towards digital construction.Utilizing advanced information technology to support urban planning,construction,operation,management and emergency response has become an effective way to improve the efficiency of urban management and promote the development of digital cities.Buildings are an important part of city composition,and three-dimensional modeling around buildings has also become a very important part of digital city construction.The method of 3D modeling of buildings is continuously developing in the direction of high efficiency and speed.As a new modern space mapping technology,3D laser scanning technology is widely used for its high efficiency,high precision,and full automation.Many areas.The 3D laser scanning technology is used to obtain the convenience of the point cloud of the building,and the3 D model of the building is constructed on the basis of the point cloud frame.However,the amount of original point cloud data acquired by the 3D laser scanner is very large,which causes great difficulties for subsequent point cloud modeling.How to streamline the original point cloud data with a huge amount of data,and extract the building point cloud framework on the basis of retaining the key information of the building point cloud has become a current hotspot of research.Based on the research of previous experts and scholars on point cloud data reduction methods,this paper proposes an adaptive point cloud reduction algorithm.Taking the library of Wuhan University of Technology Nanhu Campus as the modeling object,the point cloud data was acquired using a 3D laser scanner,the point cloud data was streamlined and a 3D building model was constructed,and the point cloud error model and error reliability index were obtained by derivation.The calculation method performs error entropy evaluation analysis on the validity of the geometric information displayed based on the threedimensional model.The main work done by the thesis includes:(1)The research status of 3D laser scanning technology,point cloud data reduction,and 3D modeling of buildings are introduced.The measurement principle,data preprocessing of the 3D laser scanning technology,and the influencing factors of point cloud accuracy are analyzed.(2)Aiming at the problems of processing the original point cloud data and the existing point cloud reduction algorithms with low operating efficiency,many noise points,and incomplete retention of key point cloud information,a new point cloud adaptive reduction algorithm was proposed.The algorithm fully considers the differences between the feature point cloud and the non-feature point cloud in the modeling framework information.The comparison results of examples show that the algorithm can maximize the streamlining of the point cloud while retaining the key point cloud information,and compare with the existing point cloud.Compared with the algorithm,the algorithm has high operation efficiency,obvious reduction effect of redundant points,and key point information retained is relatively complete.Considering that the field scanning adopts a multi-station method,a global point cloud registration algorithm is proposed,a point cloud registration model is established,and multi-site cloud data is unified under the same coordinate system based on solving registration parameters of points with the same name.The feasibility of the registration algorithm is illustrated by calculating the angle error between the three-dimensional coordinate error of the feature point and the fitted feature surface.(3)Taking the library of Wuhan University of Technology Nanhu Campus as a typical three-dimensional modeling object,field scans are used to obtain cloud data from multiple measurement sites,and point cloud registration is unified into the same coordinate system.The global point cloud is simplified,and the block modeling and texture mapping are completed based on the point cloud framework.By analyzing the laser ranging,angle measurement and registration errors,the relationship between information entropy and error entropy is deduced,and a point cloud error model is established.Based on this model,the calculation method of the error reliability index is obtained.Comparing the traditional measurement method with the geometric information difference and error reliability index based on the 3D model,the accuracy of 3D modeling is illustrated.