Texture Reconstruction Of Building Facade Based On Vehicle LIDAR System

Vehicle-Borne Laser Scanning System is a bran-new technical for Geographic information data collection and to provide a new technical for reconstruction of 3D and smart city,it becomes the focus of many scholars because of its information acquisition accurately and fast. This dissertation focuses on the theme of building facade reconstruction based on vehicle-borne LIDAR in reconstruction of 3D city. It amply introduces the research status of the development of the Vehicle-Borne Laser Scanning System and reconstruction of building facade, it also elaborates the composition and basic principle of Trimble MX2 and point cloud data’s characteristics and pretreatment. On this basis, the paper introduces the extraction of building point cloud, acquisition and process of CCD image, matches of point cloud and image. The idiographic research contents are as following:(1)It expatiates the components and principle of vehicle-borne LIDAR System, the conversion of point cloud, the acquisition of image, the characteristics of point cloud are analyzed. It amply introduces the technology of building facade texture acquisition,the issues which need to be settled are summarized.(2)On the existing grid filtering basis, a improved grid filtering which the elevation information is put in the Density of grid projection is put forward, the efficiency of the method is proved.(3)The traditional method of image to segment image by K-means clustering algorithm. Using the method based on mathematical morphology to smooth the edge and getting the image can be matched with point cloud data; The methods based on color of grid removes the shelter,for example the trees shelter.(4)It introduces texture reconstruction theory, the corresponding nexus between the image and point cloud is established using the geometry relationship of sensor and POS data of vehicle-borne LIDAR System, which is the focus of the paper. Experiments show it can be realized the texture reconstruction.