【作者】 李军；

【导师】 李永树；

【作者基本信息】 西南交通大学 ， 大地测量学与测量工程， 2012， 硕士

【摘要】 无人机遥感系统具有成本低、机动灵活、携带方便、适宜高危地区勘测等特点,对载人飞机航空遥感和卫星遥感进行了有效的补充,具有广泛的市场开发前景。在灾害应急方面,无人机具有起飞不用专门的跑道、能够有效的避开云层和天气的干扰、获取的影像分辨率高、执行任务时机动灵活等优势,相对于载人飞机航空遥感和卫星遥感具有独特的优势。本文以5.12汶川地震灾区某村镇灾后重建为试验区,总结了无人机遥感信息快速获取关键技术,研究了无人机影像的四种快速拼接方法,阐述了无人机遥感技术实时、准确地获得了地震灾区灾情信息,制作了地震灾区正射影像图和DEM,论述了无人机制作大比例尺地形图方法,探索了利用无人机遥感影像制作三维景观图并通过计算机系统进行展示。本文的主要研究内容和创新意义如下：1)从无人机遥感的参数预设定、影像重叠度、航线弯曲度和航高差等方面介绍了无人机影像质量的快速评定方法；从去云雾、匀色裁边、畸变差改正和重采样等方面介绍了无人机影像的预处理工作。2)研究了快速拼接无人机全景影像图、基于POS数据和基于SIFT算法的影像拼接方法,通过对SIFT算法计算步骤分析提出了改进方法,通过该算法的改进能够节约大量的计算时间;利用Visual C#2008编制了相关软件实现了基于SIFT算法的无人机影像自动拼接。3)在地震灾区通过无人机遥感试验,获取了灾区无人机影像数据；通过数据处理并进行误差分析,说明了无人机制作1：2000大比例尺地形图的可行性。对比分析无人机制作大比例尺地形图和其它成图方法在适应性、效率、精度和成本方面的差异,阐述了在地震灾区制作大比例尺地形图的方法并进行了制作大比例尺地形图试验。4)阐述了利用无人机遥感影像制作地震灾区三维景观图步骤,并制作了地震灾区三维景观图,该三维景观图具有分辨率高和形象逼真等特点。通过对三维景观展示系统的编制,实现了地震灾区三维景观的浏览、查询和分析功能。更多还原

【Abstract】 UAV Remote Sensing system (UAVRS) has the characteristics of low cost,flexible,easy to carry and suitable for high-risk areas surveys,an effective complement to manned aircraft airborne remote sensing and satellite remote sensing,with a wide range of market development prospects in disaster emergency. UAVRS can fly without a dedicated track.UAVRS can effectively avoid the interference of clouds and bad weather.UAVRS can obtain high image resolution and work flexibility. UAVRS has a unique advantage to compare to manned aircraft airborne remote sensing and satellite remote sensing.This article uses the reconstruction village of the5.12Wenchuan earthquake disaster area as pilot area and summarizes the key technologies of quickly access to UAVRS information and studies the way of four fast splicing with UAV images and describs accurate disaster information in earthquake disaster area with UAVRS technology. This parer produces orthophoto map and the DEM of the earthquake-stricken areas, discussing production of large-scale topographic maps with the UAV, exploring the use of UAVRS images to create three-dimensional topographic map and on display through computer systems.The main research of content and innovative as follows:1) The pater introducts the rapid assessment of the UAV image quality from such as parameters pre-set, image overlap, route curvature and the difference of flight height And introducts the UAV image preprocessing from such as Eliminate the clouds, uniform color, distortion error correction and resampling.2) Study on the image of the splicing method based on rapid splicing UAV panoramic image, POS data and SIFT algorithm. The pater Improves method by the SIFT algorithm computing step analysis and saves a lot of computing time through the improvement of the algorithm. The preparation of the related software with Visual C#2008achieves the UAV images automatically stitching based on SIFT algorithm.3) By the UAVRS experiment in the earthquake area, we access to the disaster area of UAV image data.The data processing and error analysis show the feasibility of UAV production1:2000scale topographic maps. The paper analysis the differences of the UAV production of large scale topographic maps and other mapping methods in adaptability, efficiency, accuracy and cost,describing the method of large scale topographic maps produced in the earthquake area and the production of large-scale topographic map test.4) Elaborating the step of producing three-dimensional landscape of the earthquake disaster area with UAVRS images and produced a three-dimensional landscape plan for earthquake-stricken areas. The three-dimensional landscape plan has high resolution and vivid image. Through the preparation of three-dimensional landscape display system,the system has the capabilities of browsing, query and analysis obout the three-dimensional landscape of theearthquake-stricken areas.更多还原