Based On The Lidar Points Clouds Building Airborne3D Modeling Technology Research

Light Detection and Ranging is a kind of new earth observation system of beingwidely used application and research,It has hightly accurate of3D space information andthe DSM data, which enhances the accuracy and speed of survey and automate acquisitionand processing of the data. The LIDAR provides a new observation techniques in theprocess of digitalization of accelerating today for a3D model of digital city.But because ofthe city’s buildings in3D reconstruction of process modeling algorithms was lack ofautomation, greatly limits the LIDAR technique in building the application of modeling.In3D city construction process, how to realize the building of the model and theextraction of3d reconstruction has been the focus of research, this article takes theairborne laser radar data system as the research object,the paper analyses the characteristicsof laser radar point cloud data and data filtering, and on the basis of research on the citybuilding houses contour extraction, feature point detection and3d reconstruction work, thearticle main research work are as follows:1) Through to the airborne laser radar system components of the system and workingprinciple on the basis of the analysis of, summarized the airborne LIDAR and city building3d reconstruction, the research present situation and related technical theory, and pointedout the at present in this field of deficiency and difficulty.2) Analyses laser point cloud data characteristics and organizational form, this paperintroduces the ASPRS proposed LAS file standard file format, discusses the organizationform of several common data, using the data-trees KD index structure to the huge numberof laser point cloud data for a rapid index, to improve data processing efficiency and theprecision of3d reconstruction.3) The building housing the outline of the outline of the boundary line featureextraction, in contrast to the existing research of the algorithm based on the2D Houghoperator to realize the building characteristic line extraction, and through the outlierclassification, line and merger right Angle equation constraints means to features of theoptimization of the line. In the test the contour line by random sampling after consistency (RANSA) operators to realize the generation of building roof plane detection, againthrough the LIDAR data and image registration to realize the generation of roofmathematically processing, the last through the problem of the way to realize the mapbuilding three dimensional model of the distinct, visualization.