|Alternative Title||Study on Debris Flow Three Dimensional Virtual Reconstruction Based on Remote Sensing Data|
|Place of Conferral||北京|
|Keyword||泥石流灾害 遥感图像几何配准 遥感图像分类方法 泥石流动力过程模型 三维虚拟再现|
|Abstract||泥石流灾害是我国西南山区一种较为常见的突发性自然灾害。在泥石流灾害发生后，快速获取灾情信息，形成救灾决策，部署应急救援对减轻灾害的影响有重要意义。遥感技术的发展为实现这一目标提供了技术手段。遥感技术应用于泥石流灾害研究，虽然已经取得了许多成功的经验，但是大多停留在应用单时相遥感图像识别或灾害前后图像对比分析等方法解译灾害信息。由于泥石流灾害是一种突发性地质灾害，发生过程时间短，对泥石流的运动过程了解，有利于准确认知泥石流的流经区域和致灾形式与程度。因此，本文以无人机遥感图像和卫星遥感图像为数据源，提出了基于遥感数据的泥石流灾害的三维虚拟再现的方法，综合遥感技术、泥石流动力过程模拟技术等技术，通过遥感技术提取泥石流灾害灾情信息，虚拟重建泥石流灾区三维环境，并通过简化的泥石流运动模型，虚拟再现了泥石流灾害的发展过程。该方法能够在灾害发生后迅速获取遥感数据，并通过虚拟再现技术弥补了传统泥石流遥感调查的不足。在研究工作中取得了以下成果： (1) 探讨了泥石流三维虚拟再现中所用到的关键遥感技术。灾害发生后，无人机遥感技术能够及时提供灾区现场遥感数据，结合灾前卫星遥感图像，能够对灾区进行较为详细的遥感分析。本文首先针对泥石流流经区域的特点，提出了基于线性特征的遥感图像几何精配准算法。然后，利用SPOT5图像结合数字高程模型和坡度、坡度等数据，提取泥石流流经区域的分区算法；最后，利用面向对象的分类方法处理无人机遥感图像，提取灾区建筑物空间分布信息。 (2) 推导并简化了基于深度平均理论的泥石流动力过程模型。稳定、可靠的泥石流动力过程模型是泥石流三维虚拟再现的核心。本文基于连续流体假设，基于深度平均理论推导了泥石流动力过程模型方程，并对模型的可靠性进行了详细分析。然后，采用有限体积数值离散方法在DEM规则格网上对泥石流动力模型方程进行数值离散，经过算例分析，证明本方法是稳定和有效的。 (3) 应用已经建立的模型，设计和实现泥石流三维虚拟再现系统软件，在虚拟环境中对不同条件下泥石流动力过程进行虚拟实验研究。通过试验，讨论了不同条件对泥石流动力过程的影响，研究其规律和特征，同时论证系统的可靠性。 (4) 最后，以实际泥石流为例，虚拟再现其动力发展的过程，并将虚拟再现所得的泥石流堆积扇范围及掩埋居民建筑面积等信息，与实际泥石流灾害记录进行对比分析，结果具有较高的精度，论证本文研发的泥石流三维虚拟再现系统在实际应用中的可靠性。|
|Other Abstract||Debris flow is a common natural disaster in mountain area of Southwest China, and is a|
serious threat to the local people's life and property safety. After the debris flow disaster, it is important significance to reduce the impact of disasters to collect information rapidly, form a reasonable relief strategy and design a relief scheme. Timely, accurate disaster information is the prerequisite and guarantee of the work of emergency relief. The development of Remote Sensing (RS) technology provides technical means to achieve this goal, especially the rapid development of high-resolution remote sensing technology, the efficiency of access information is more quickly, the number of access information is more abundant. The researches, which use high-resolution remote sensing technology in debris flow work, have provided rich experience, but most tests is focus on the application of single phase remote sensing image to obtain disaster information. As the debris flow disaster is a sudden geological disaster, occurred in a short time, the traditional method is difficult to describe the motion process of debris flow from remote sensing images. The understanding of the process of debris flow motion, we can accurately cognitive flowing through the area of the debris flow and lead to the causes of disasters. So, we present a three-dimensional virtual reproduction method of debris flow which based on the remote sensing data in this article. And the method integrate remote sensing technology and mudslides dynamic process simulation technology and other technical, through remote sensing technology to extract the debris flow disaster-stricken area, reconstruct a three-dimensional scene in a virtual environment, and by simplifying the mudslides dynamic process model, we simulated the process of a debris flow
event in a virtual environment. Following results were obtained in the research work.
(1) Explore the key remote sensing technology used in the debris flow three-dimensional
virtual reproduction. Firstly, we analyzed the characteristics of debris flow basin remote
sensing data, according to the characteristics of there is less control points in the debris flow basin remote sensing image, we propose a line based image correction method to process remote sensing image. Then, according to the characteristics of the debris flow, we use a multi-source remote sensing data to extract debris flow information. For example, we use the SPOT5 images combined with digital elevation model and Slope and Aspect data to extract different area of debris flow basin with the support vector machine classification method, and achieved satisfactory results; we used object-riented classification approach to process UAV remote sensing image, extract the important of buildings' spatial distribution information, demonstrate that the high-resolution remote sensing images is useful in extract buildings information.
(2) We derive a debris flow dynamic process model based on the depth of the average
theoretical, and simplify the model. Stable and reliable model of debris flow dynamic process is the core of the debris flow three-dimensional virtual reproduction. This article is based on the assumption of continuous fluid and the depth of the average theoretical, and then derives the debris flow dynamic process model equations. And in the process of model derivation, in order to improve the computing performance, we reasonably simplified model, and analyzed the reliability of the model. Finally, based on the approximate Rimann solution of HLL format, we discrete the DEM grid rules debris flow dynamic model equations numerically using finite volume discretization method. After the numerical analysis, we have proved that the method is stable and effective.
(3) We applied the model has been established, in a virtual environment, and we carried
on Virtual Experiment debris flow dynamic process test under different conditions. Verified by experiment, we discussed the different conditions of debris flow dynamic process, and study the rules and characters. Including terrain conditions, at the end of resistance conditions, obstructions and other dynamic processes.
(4) Finally, take Wenjiagou debris flow for example, virtual reproduction the process of
debris flow disaster, and comparative and analysis the calculated hazard range and the actual situation. After the analysis of the dynamic process phenomena and problems, we get a conclusion, the debris flow three-dimensional virtual reproduction technology is feasible in practical applications.
|臧文乾. 基于遥感数据的泥石流三维虚拟再现关键技术研究[D]. 北京. 中国科学院研究生院,2013.|
|Files in This Item:|
|臧文乾：基于遥感数据的泥石流三维虚拟再现（19369KB）||开放获取||CC BY-NC-SA||View Application Full Text|
|Recommend this item|
|Export to Endnote|
|Similar articles in Google Scholar|
|Similar articles in Baidu academic|
|Similar articles in Bing Scholar|
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.