【作者】 马瑞升；

【导师】 黄耀； 孙涵；

【作者基本信息】 南京农业大学 ， 环境工程， 2004， 硕士

【摘要】 本研究针对现有遥感技术存在的影像资料实时获取能力差、获取成本高的问题而展开。在借鉴国内外无人机应用经验的基础上，提出了将无人机与数码相机结合构成微型无人机航空遥感系统，实现影像获取的解决方案。研究工作主要分为无人机遥感系统的集成、改进、飞行作业与无人机遥感影像的几何纠正两方面内容。 在系统集成阶段，对数码相机在航空遥感中的应用作了分析，指出现有数码相机中的高档机型已能满足大多数遥感应用的要求。在对系统的技术改造中，测定了Ni-MH电池组放电曲线，对无人机机载电源做了改进，配合加装的微波视频传输系统和自行研制的降落阻尼拦阻网可以大大提高无人机航空遥感系统的安全性、适应性和作业能力。在介绍系统的日常飞行作业程序后，给出了广西武鸣县飞行任务实例，为系统的推广应用提供了借鉴的依据。系统的飞行试验证明无人机遥感系统具有机动灵活、反应迅速、适应性强、获取的影像分辨率高和使用成本低等诸多优势，是卫星遥感和航空摄影的有益补充。 微型无人机遥感由于受到无人机平台体积小、重量轻、任务载荷有限等因素的制约，在飞行作业中传感器位置和姿态受气流影响大、稳定性差，而且目前是非专业摄影机成像，这些都大大增加了其影像几何纠正处理的复杂性。通过全面分析遥感影像几何变形的来源和现有几何纠正方法，提出了一条基于CCD相机误差综合标定和与正射影像配准的几何纠正技术路线。 通过对现有CCD相机标定方法的综合分析，提出利用建筑物立面现成规则图形纹理结构的垂直摄影法来完成CCD相机综合误差的标定。该方法不使用特殊仪器，也无需人为布置大量控制点，只需做简单的标记和量测工作就能完成标定影像的拍摄。将获取的影像与数学模型结合即可快速、简单、有效地解决CCD相机综合标定问题。 借鉴卫星影像的处理方法，将目标区现有的大比例尺地形图数字化后，利用多项式法完成了广西武鸣城区无人机遥感飞行试验影像的几何纠正，并取得了较好的效果。经检验，单点定位平均偏差为1m左右，最大偏差为2m左右，在X、Y方向上的长度变形均不大于1％，基本没有角度变形。作为基准影像，大比例尺地形图在自身成图精度和所含特征信息等方面有一定的局限性，如使用高分辨率卫星影像作为基准底图进行配准纠正，则成图精度会有进一步提高。徽型无人移以晓空遥感系统及其影像几何纠正研究 本研究结果证呢，将普通数码湘机和无人机结合构成微型无人机航空遥感系统，在技术上是可行的，其获取的高分辫率遥感影像经几何纤正处理后可与载人航空摄彩、国外高分辫率成像卫星岭影像资料相墉美，是一项低成本、快逮解决小区城可见光、红外等遥感信息源问题的可行技术.更多还原

【Abstract】 Now, there are two problems in Remote Sensing (RS) techniques: images can not be obtained in real time and it is very expensive to obtain by plane or satellite. Aiming at these situations, a miniature airborne optical RS system which consists of a Digital Camera (DC), a miniature robot plane(Unmanned Aerial Vehicle, UAV) and its’ ground-based subsystem was presented on the basis of analyses of the applied experiences of UAVs abroad and home. In this paper, two type problems about this miniature RS system would be discussed: one was system integration, amelioration and flight planning; another was geometrical correction of images from this system.In the stage of system integrating, the applications of DC in aerial RS was analyzed, the conclusion was that the advanced models in actual digital cameras had been satisfied to most RS application. In technique rebuilding of system, discharge curve of Ni-MH battery was measured, and power supplies in UAV were improved. The security, adaptability and efficiency of miniature aerial RS systems could be improved obviously by microwave video transmission system and self-regulating blocking net in landing. As a pattern, the flow of flight planning of WuMing county in GuangXi province was introduced, it was good to the planning of similar RS flight. The flight experiments had proved that the miniature aerial RS system based on UAV would be the useful supplement of satellite RS and aerial photography. The advantages of this system are as follows: obvious flexibility, broad adaptability, rapid response, low cost and the ability of acquiring high-resolution images.Because UAV is smaller and lighter than normal plane, the miniature aerial RS system is easily effected by wind, its’ stabilities of position and stance are not good. Because the load of UAV is so finite, the large professional aerial photogrammetry camera can not be installed. Considered as size, weight and price, the Charge Coupled Device (CCD) DC is the best choice for the UAV RS system. The factors mentioned above would bring lots of complicated geometrical distortion to the images from the UAV RS system. On the basis of analyzing roundly origins of RS image geometrical distortion and existing methods of geometrical correction, the route of geometrical correction for the images from theminiature UAV RS system was brought forward. This technical route was based on integrated CCD camera calibration and image registration by orthograph.CCD camera calibration was accomplished by vertical photograph on the wall of buildings. This method was aimed at the texture of existing figures on the wall, so it did not need use special instruments and array lots of reference points, it only needed some simple marks and measuring work to accomplish the photograph of base image. The problems of CCD camera calibration could be solved effectively and simply by the obtained image and mathematic model.By means of images processing of satellite, geometrical correction of images from UAV RS system was accomplished by polynomial. In the processing of images of WuMing in GuangXi province, a set of scan large-scale relief maps were used. By testing the corrected image map, the mean error of point position was 1 meter, the maximal error was 2 meters, length distortions on X and Y axis orientation were not more than 1% and angle distortion was almost not existing. As the basic image, large-scale relief maps were not good, with some disadvantages of accuracy and surface features. If a high-resolution satellite image could be used to registration as a basic map, the accuracy of corrected image map would be improved obviously.This study has proved that it was feasible that common DC and UAV were combined to construct a miniature aerial RS system. After geometrical correcting, the high-resolution RS images from the UAV RS system were as good as the images of satellite or aerial photogrammetry. The miniature UAV RS system is a new technique that could be used to quickly acquire images of small region with low-costs.更多还原