Study Of Buildings After Major Earthquakes Based On Terrestrial Three-dimensional Laser Scanning

Earthquakes can cause mass of people death and property damage, for which the collapse of buildings is one of the main reasons. Thus, investigation of characteristics of building damage is of great importance for research of seismic disasters. The traditional way dominated by direct observations can only yield limited information, which cannot record more comprehensive scenes of affected areas, so posing limitation to in-depth analysis of building damage. Fortunately, the emergence of new technologies brought the possibility of quantitative research to seismic hazard. For example, application of remote sensing has achieved good results in the post-earthquake damage assessment, and provided an important basis for rapid damage assessment. Earthquakes are, however, essentially of a random process. After a major temblor occurs, the timely acquisition of remote sensing data for an accurate assessment of earthquake damage plays an important role in post-seismic assessment. Along with the emergence of airborne LiDAR, it is possible to acquire image data after an earthquake at the first time, and its high accuracy for post-earthquake building damage assessment provides reliable information. But due to the airborne LiDAR acquisition point interval limit and precision defect, as well as small deformation of buildings, it cannot accurately reflect the deformation features of houses after an earthquake. The terrestrial three-dimensional laser scanning can overcome the above defects. Compared with conventional technology, this new technique has many advantages such as high precision, fast speed, true reflection of the prototype, quick extracting relevant information, and quantitative analysis with high-precision, high-density, and non-contact. The results show that the terrestrial three-dimensional laser scanning technology provides a new approach to investigate quake-caused building damage.Using Trimble GX200, this work has collected LiDAR data of 351 houses in the Lushan region after the April 20,2013 Lushan M7 earthquake. The data processing was completed in MATLAB 2010. The idea of this method is that the a house is divided into individual walls, which are transformed to the wall coordinates, i.e. the plane perpendicular to the direction of the wall to represent the deformation of the wall. Such work was also made after the August 3,2014 earthquake in the Ludian region Yunnan Province, where 325 houses data were collected, and the data processing method proposed in this thesis was used to handle these houses data. The results show that this method is suitable for building point cloud data processing. With its high accuracy it can reflect housing millimeter deformation. From data processing, the resulting deformation maps correspond with the photos of houses. But the deformation features such as tilts, cracks, and distortions are not directly related with housing damage levels. The house processing point cloud method of this thesis is applicable to the houses with regular structure, not for irregular buildings, timber frame and wooden structure. The analysis of five different types of structures with varied damage grades house yields the following results:For those buildings with negligible to slight damage, deformation of-lcm-lcom occurs at some places prone to uneven change, such as ring beams, structural columns and windows, and house edge. There are also some areas with deformation reaching 2-3cm. For the houses suffered from moderate damages or more, the deformation at apparent cracks amounts to 2-3cm, some are even smaller. It indicates that a smaller amount of deformation caused by the earthquake. Besides, this thesis also suggests some methods used for analysis of house point cloud, such as histograms and contour maps.