Research On Buildings’ Damage Measuring Methods Based On Vehicle-borne LIDAR Data

One of the most urgent task, that when serious natural disaster happening, is to acquire disaster information of buildings. However, heavy weather accompanying with disaster restraints the quick information acquisition about condition of disaster,so the initial period is the most lack of information period. Vehicle borne LiDAR system has advantage forbuilding disaster information acquisition, it can arrive disaster area in time when disaster happening, and set to work at once. This project aimed to solve the question of emergency requirement of burst nature disaster, based on disaster environmental data acquired by vehicle-borne LiDAR system, research the methods of extracting building damage information from point clouds. The paper selected a number of buildings and their scaling models as research object, using some methods such as experimental simulation, comparative analysis and data processing to research the methods of vehicle borne LiDAR serving for building damage emergency measure,which gets along in theory and arithmetic to a certain extent.The major works implemented in this dissertation are listed as follow:1.Segmentation the building point cloud in city scene based on use existing geographical based date as auxiliary data. Firstly convert the coordinate system of fundamental geographic data to that of vehicle borne LiDAR point cloud, and realize the coordinate system between them; secondly extract effective contour information of the buildings from the existing geographical based date;thirdly, set a suitable buffer threshold value for building contour, and segment out point cloud data of building facades automatically.2. The aspect is on the research of use point cloud density and elevation for the fine filter. Dividing into grids based on the range of original point cloud,setting the threshold of grid number and grid average elevation, the grid which is meets the threshold conditions is remained,the buildings are extracted from point cloud data.3.Making the building disaster measurement simulation test.Different buildings’ three-dimensional models after the disaster are constructed.Making a scaling model to simulate the building,construct several main building damage models,and acquired equivalent point clouds corresponded to buildings actual damage.4.Segmentation of building facade point cloud based on Random Sampling Consensus is implemented and extractation the characteristic line based on convex hull algorithm.In the segmentation devise segmentation judgement rule based on the point cloud plane equation and the r-radicus density in Computer Graphics, the segmentation flow is presented.5.Summarizeing the morphological characteristics of the buildings under different damage forms based on extracting building damage information,estimating damage style and damage degree, providing new technical support for serious natural disaster emergency.