Extraction Of Three-dimensional Informationof Buildin Gs And Seismic Damage Forecasting Research Based On High-resolution Remote Sensing Images

Earthquake is one of the serious natural disasters faced by all humanbeings. With the rapid development of the modern society, the urbanpopulation and the number of buildings are more and more. Once thedestructive earthquake occurs, their damages are especially severe. Therefore,good seismic damage forecasting of city is of a very important practicalsignificance. With a regard to the researches on seismic damage forecasting ofcity for a long time, three-dimensional attribute information of the buildingsare mainly acquired by manual investigation, which takes a long cycle andhigh costs and is hard to keep pace with the development speed of the moderncities, thus losing the effectiveness of data. With the fast evolvement ofremote sensing technology, researches on high-resolution remote sensingimages are gradually increasing in the aspects of seismic damageinvestigation and seismic damage forecasting. However, building height,building area, the number of buildings and other detailed building data arestill missing currently in Inner Mongolia area, and there are few studiesadopting high-resolution remote sensing images to extract buildinginformation, which affect the detailed seismic damage forecasting work inthis area. With the purpose to acquire basic building information of highaccuracy and its application, taking the urban Saihan District, Hohhot as theexample, this paper studies the method to extract three-dimensional attributeinformation of the buildings from high-resolution remote sensing images, andverifies the applicability and accuracy.By analyzing the characteristics of remote sensing image in the study area,the objects (buildings) are clearly visible in high-resolution remote sensing images. However, profiles of building roofs are diversified and spectralcharacteristics of roofs are inconsistent due to different making materials.Building shadows are visible in the high-resolution remote sensing imageseven more clearly, the spectral information of the shadow part is uniform to acertain degree, and the gray value is significantly lower than that of othersurface features. According to above features, the number of buildings andbuilding areas are extracted to apply the visual interpretation method, so as toextract the shadow information of buildings from the regional high-resolutionremote sensing images, and predict the building height. By means of fieldinvestigation and measurement, the practical and accuracy of the extracteddata are validated. Finally, earthquake with the maximum influence degree onthe Wusutu-kuisu fault in Daqingshan piedmont fracture is evaluated in termsof the extent of building damages, economic losses and casualties by takingadvantage of the acquired building data. This study in order to master thisfracture earthquake impact on the biggest influence in the study area and theminimum degree, calculate the furthest in the earthquake epicenter and therecent earthquake intensity values.As a result: When the earthquake intensityof VIII degree, the total area of32140818㎡building damaged in the studyarea, House number is6173and building its own destruction causedeconomic losses of about94billion yuan and number of casualties of up to2729during the day and4818in the night. When the earthquake happened inthe recent earthquake intensity for IX degrees, the total area of41470685㎡building damaged in the study area, House number is7295and building itsown destruction caused economic losses of about184billion yuan andnumber of casualties of up to6822during the day and11301in the night.Above extraction method of three-dimensional building information andresearch objects can provide a reference basis for the seismic prevention anddisaster mitigation of Inner Mongolia area, having some practical value.