Three-dimensional Modeling Of City Buildings Based On Airborne LiDAR And Oblique Photography

City is the crystallization of human civilization, urbanization is an important content of economic and social modernization.Urban informatization has become a new opportunity of urban development in our country, and was named "digital city" and "wisdom city" reputation. "Digital city" gradually become a new methods of management, city planning, the future cities about the construction of "digital city" is becoming more and more urgent.As the digital city construction of three-dimensional city model, one of the core content of geographic space frame data, is developed from exploring stage to depth and breadth.Building as the main element in the city, for its high efficiency, high precision of 3D reconstruct is the key to the construction of "digital city", the study of urban building modeling technique is particularly important.Airborne LiDAR technology is carried by the plane, use the laser as medium,quick access to get high precision 3D point coordinates of the ground in a large area of the feature;Tilt photogrammetry technology, through the flying platform, carry multiple sensors, gain ground features images from multiple perspectives.By combination of two new measurement technology, this paper expounds the modeling method which based on airborne LiDAR and oblique photography, based on the direction search algorithm is proposed for building contour extraction quickly,supplemented by height information, achieve the purpose of fast and efficient construction building 3D model.In this paper, the main research content is as follows:(1) the existing modeling techniques are introduced and analyzed, summed up the deficiency of the existing modeling methods and shortcomings.On this basis,expounds way of modeling which based on airborne LiDAR and oblique photography,the purpose to solve the problem of the existing modeling methods, improve the efficiency of modeling, and quality.(2) the point cloud filtering and classification algorithm was studied, this paper introduces the current several more commonly used filtering and classification algorithm, and describes the iterative filtering algorithm of triangulation encryption and area ratio of echo combined with topology identification model of classificationalgorithm.(3) This paper put forward based on the direction search building contour line extraction algorithm quickly, quickly extract the contour line, and improve modeling efficiency of the building.The algorithm the foot laser projection rules of buildings to the two-dimensional planar four polar point obtained as datum of segmentation,search, will be building boundary is divided into four segments of the search, and with a search of the four reference point.In boundary point relative datum of azimuth maximum in the start search direction corresponding quadrants determine benchmark,search out of the first boundary point, and to search the boundary point for the new benchmark continue to search.To the new reference point search and search for the next point, by calculating the new reference point and a reference point and search point Angle, take the counterclockwise Angle corresponding to the point for the next largest boundary point or turning point.Repeated the process, the search to the extreme value point, to complete each segment search, to extract the contour line building.Experiments show that the algorithm can achieve the purpose of rapid extraction buildings contour.(4) the utilization of airborne LiDAR point cloud data and images, modeling of urban buildings.Airborne LiDAR point cloud data, after filtering, classification, laser point feet, isolated buildings by this algorithm is used to extract contour line,supplemented by height information, a preliminary white building model is set up.Complete fine after editing in 3dmax, tilted by image processing platform, will tilt texture mapping to the buildings on the white model, after the building in 3dmax model integration.The experimental results show that this kind of modeling method has achieve the purpose of improve modeling efficiency and quality.