【作者】 王聪华；

【导师】 林宗坚； 陈天恩； 孙杰；

【作者基本信息】 山东科技大学 ， 大地测量学与测量工程， 2006， 博士

【摘要】 无人飞行器(Unmanned Aerial Vehicle-UAV)又称无人驾驶飞行器，我们这里主要是指无人驾驶固定翼型飞机、无人驾驶直升机和无人驾驶飞艇，采用无人飞行器作为遥感平台进行低空监测和摄影的系统我们称它为无人飞行器低空遥感系统(UAV Low Altitude Remote Sensing System，简称UAVRS系统)。UAVRS系统所获取的影像不同于传统的航空摄影影像，影像存在像幅较小(Nikon D100相机的像幅为23.7mm×15.6mm，Canon 5D相机的像幅为35.8mm×23.9mm)、数量多(一条航线最多时20多张影像)、影像的倾角过大(根据统计κ角在±35度之间)、重叠度不规则(存在航向重叠度≤30％，旁向重叠度≤20％)等问题，这些问题的存在给影像匹配、影像的定向和空中三角测量等内业处理都带来了一系列困难，而这些问题，都是国内外现有的数字摄影测量系统面临的新问题，解决这些问题才能真正体现UAVRS系统的特点和价值所在。 为解决UAVRS影像数据处理存在的问题，论文主要在以下方面展开研究： (1) 分析UAVRS影像系统误差的来源，给出畸变差改正的数学模型，进行畸变差改正。 (2) 研究UAVRS影像匹配的方法及如何提高大倾角、不同比例尺和不同灰度影像匹配的精度和有效性，探讨剔除粗差的办法。 (3) 研究UAVRS影像单模型建立的问题，如何提高单模型相对定向和绝对定向的精度，如何解决前方交会精度差的问题。 (4) UAVRS影像的空中三角测量研究，探讨UAVRS影像区域网构网的方法，给出带权观测值光束法区域网平差的数学模型，通过UAVRS影像空中三角测量试验，提出一套UAVRS影像数据处理的作业方法、作业技术流程和作业规范，探讨外业对内业数据处理影响的因素，提出改进办法。 (5) 设计一套UAVRS影像数据处理实验系统，进行地形图立体测图试验，探讨UAVRS影像经摄影测量处理后进行立体测图的可行性以及能达到的精度，进行UAVRS影像DEM生成试验，研究正射影像图生成方法和作业规范以及正射影像能达到的精度。更多还原

【Abstract】 The Unmanned Aerial Vehicle-UAV is also referred to the Unman-droved Aerial Vehicle that we mean the unman-droved plane with the fixed wings, unman-droved helicopter and unman-droved airship. We named the system making the low altitude measurement and photography through the UAV platform as the UAV Low Altitude Remote Sensing System, simply as UAVRS. The image which is acquired by UAVRS system is different from image of aerial photography on tradition. The captured image has problems such like small format images (Nikon D100 camera’s format image are 23.7mm×15.6mm, Canon 5D camera’s format image are 35.8mm×23.9mm) , excessive quantities of images (at most more than 20 images per route) , over tilt angle of images (as statistic k between ± 35) , irregular overlap (exist problems like forward overlap ≤30% , side overlap ≤20%) etc. These problems exist brings a whole series of trouble into image matching, image orientation and aerotriangulation and some other relative subject. It is also the new problem of the Digital Photogrammetric System that both national and international are facing. Only the solving of these issues can reflect the specialization and value of the UAVRS.To solve the existing UAVRS image data processing problem, the dissertation has studied these following subjects:(1)Analysis the source of image system error, give the correct mathematics model of distortion.(2)Study the methods of UAVRS image matching, how to improve the accuracy and validity of different scale and different grey level image matching and over tilt angle of images , discuss the solution of eliminating gross error.(3)Study the UAVRS image single model build problem. How to improve the accuracy of single model relative orientation and absolute orientation. How to improve the accuracy of forward intersection.(4)Study the UAVRS image of aerotriangulation, discuss the methods of build UAVRS image block . Give an operation methods, operation technique procdure and standard operation regulation. Discuss the effective element that made by the fieldwork to offices data更多还原