无人机影像在地震灾区道路损毁应急评估中的应用研究

发布时间：2018-03-23 21:37

本文选题：无人机　切入点：应急灾害　出处：《西南交通大学》2013年硕士论文

【摘要】：汶川地震发生后,灾区地面交通遭到破坏,阻碍了救援队伍进入灾区,使本可以通过及时救助还可以生存的伤员失去被救助的机会。因此,道路交通线的快速抢通决定了救援工作能否高效地开展。如果在震后应急期间可以对道路损毁情况快速评估,决策者根据评估结果制定相应的政策使救援工作顺利开展,对抗震救灾具有十分重要的意义。在5·12救灾工作中,因受阴雨云雾天气和卫星重访周期交叉因素的制约,遥感卫星和常规航空摄影飞机无法及时获取灾区影像,而无人机系统凭借其机动灵活、云下作业的特性及时获取灾区遥感影像。获取灾区影像,只是初步工作,如何在震后应急期间,利用现有资料对畸变大、数量多的无人机影像快速处理,提取震害信息,为救灾决策者提供科学依据,是亟待解决的问题。本文以安县茶坪乡B35县道的震后无人机影像和震前DEM为实验数据,首先,借助摄影测量软件采用航带法自由网平差对无人机影像快速处理,得到正射影像镶嵌图能够为茶坪乡的灾情提供及时有效的宏观信息,满足一定的灾情监测和评估需求,但本文的道路损毁度评估是基于DEM的空间分析,需要具有实际高程信息的DEM。因此,从Google Earth影像上采取控制点,在自由网平差的基础上进行绝对定向,得到震后的DEM及DOM；对正射影像镶嵌图目视解译判别灾害体和道路并勾绘,通过ArcGIS的叠加分析功能得到道路损毁区的范围,获取各损毁路段的长度及道路总长度；通过OpenCV和GDAL实现基于特征提取的DEM自适应匹配算法,将震前与震后的DEM进行无控制点匹配；以道路损毁区的范围和震后DEM为数据源,利用ArcGIS的空间分析功能提取道路损毁区的DEM数据；然后,通过C++语言编写程序对损毁区各路段震前与震后的DEM范围,基于优化的设定阈值细分三棱柱的体积法实现土方量的计算；最后,以道路总长度、损毁长度、掩埋体的成分为因素,对损毁比例、损毁规模、受损系数进行量化,形成损毁度的评价指标；以土方量、掩埋体的成分与机械性能为因素,对抢险工期进行估算,为救灾决策者提供定量数据,对抢险救灾的顺利开展具有重要的理论和实际意义。
[Abstract]:After the Wenchuan earthquake, ground transportation in the disaster area was damaged, which prevented the rescue teams from entering the disaster area, so that the wounded who could have survived through timely rescue lost the opportunity to be rescued. The rapid grabbing of the road traffic line determines whether the rescue work can be carried out efficiently. If the road damage can be assessed quickly during the post-earthquake emergency period, the decision makers will formulate corresponding policies according to the assessment results to enable the rescue work to proceed smoothly. It is of great significance for earthquake relief. In 5 / 12 disaster relief work, owing to the intersecting factors of cloudy, rainy and fog weather and satellite re-visit cycle, remote sensing satellites and conventional aerial photography aircraft were unable to obtain images of disaster areas in time, and the UAV system was flexible by virtue of its mobility. The characteristics of cloud operations in time to obtain remote sensing images of disaster areas. To obtain images of disaster areas is only a preliminary work. How to use existing data to quickly process large and large amount of UAV images and extract earthquake damage information during the post-earthquake emergency period, It is an urgent problem to provide scientific basis for disaster relief decision makers. In this paper, the post-earthquake UAV images and pre-earthquake DEM images of B35 County Road in Chapingxiang, an County, are taken as experimental data. Firstly, the UAV images are processed quickly by using the aerial belt free net adjustment with the help of photogrammetry software. The orthophoto mosaic can provide timely and effective macroscopical information for the disaster situation in Chapingxiang and meet the needs of disaster monitoring and assessment. However, the road damage degree assessment in this paper is based on the spatial analysis of DEM. Therefore, the control point is taken from Google Earth image, and the absolute orientation is carried out on the basis of free net adjustment to obtain DEM and DOM after earthquake, and the orthophoto mosaic image is visually interpreted to distinguish the disaster body and road, and the road is drawn. Through the superposition analysis function of ArcGIS, the range of road damage area is obtained, and the length of each damaged road and the total length of road are obtained. The DEM adaptive matching algorithm based on feature extraction is implemented by OpenCV and GDAL. The DEM before and after the earthquake is matched without control points, and the DEM data of the damaged road area is extracted by using the spatial analysis function of ArcGIS, taking the range of the damaged road area and the DEM after the earthquake as the data source. The volume method of subdividing triangular prism based on optimized threshold is used to calculate the earthwork volume of each section of damaged area before and after the earthquake, and finally, the total length of road, damage length, damage length are used to calculate the volume of earthwork. The damage ratio, damage scale and damage coefficient are quantified to form the evaluation index of damage degree, and the earthwork quantity, the composition and mechanical properties of the burial body are taken as the factors to estimate the period of emergency. It is of great theoretical and practical significance to provide quantitative data for disaster relief decision makers.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237;P315.9

【参考文献】