Research On Reconstruction Of 3D Building Model Based On Laser Point Cloud Data

Three-dimension terrestrial laser scanning technology is the latest one in measurement field,which can completely obtain the point cloud data on the surface of the object.It is used widely in many fields,because it is high precision,non-contact,digital and automatic.Along with the development of computer technology and deeply research,3D reconstruct io n technology is more and more mature,the point cloud data of object is collected by 3D laser scanner,and its modeling has become a hot research topic in the present,especially for building modeling is becoming the focus of research in many fields.However,3D reconstruction technology of build ings is still in the stage of rapid development,the technology of point cloud surface reconstruction about large buildings is relatively backward,so many technical problems need to be solved about how to build 3D Buildings Model quickly and accurately,for example,how to solve the model reconstruct io n of complex detailed structure of buildings.This paper studies from the two aspects such as point cloud data processing and model reconstruction.In the progress of point cloud data processing,In view of the problem that noise points can't be effectively deleted by the denoising algorithm of eigenvalue plane fitting,because the gross error or outliers are not considered,This paper presents an improved robust denoising algorithm of eigenvalue plane fitting,The algorithm can make up for the shortcomings of above algorithm,which can effectively delete the noise points,lay the foundation for building 3D Buildings Model.In the progress of 3D Buildings Model,In view of the problem that degree of automation is not high,inefficient about the traditional methods of model reconstruction,for example,parametric model reconstruction method based on feature extraction,this paper presents a triangle mesh reconstruct io n algorithm based on the plane feature of buildings,the algorithm realizes that automatic 3D model construction,obtains good modeling results,improves the modeling efficiency,shortens the modeling time.Taking the campus buildings of Jiangxi University of Science and Technology as the experimental objects,the algorithms proposed in this paper are verified by examples,good results are obtained.Compared with the traditional denoising algorithm of plane fitting,the denoising algorit hm in this paper improves denoising accuracy,meets the modeling requirements;the model reconstruct io n algorithm in this paper improves the modeling efficiency and realizes the building automatio n modeling.The results show that the algorithms are feasible and efficient.