低空摄影测量中的北斗辅助空中三角测量方法

Analytical aerotrangulation is the best method for precise image orientation and high accurate object positioning. However, for the low-altitude photogrammetry, big image distortion, large angle of image inclination, non-standard image overlap, large photo number and large level of route bending, a large number of ground control points are necessary when performing traditional aerotriangulation and the automation level is relatively low. Based on the research on GPS-supported aerotrangulation for many years, this project firstly proposes to introduce Beidou dynamic positioning technology into the field of low-altitude photogrammetry, and intends to comprehensively and further study the precise point positioning model of Beidou satellite navigation system, the onboard resolution of phase integer ambiguity, systematic error compensation for multi-sensor integration, automatic measurement of unconventional aerial photographs, rigorous mathematical model of Beidou-supported bundle block adjustment and design of the optimized block, and intends to develop a corresponding software system. Finally, this project will raise the exposure station accuracy acquired by Beidou and GPS and then provide theoretical framework and practical scheme of replacing GPS by Beidou, and raise the object positioning accuracy of low-altitude photogrammetry to pixel level using four ground control points around the corner of an adjusted block, expecting to take advantage of the superiority of low-altitude photogrammetry in emergent survey and lage-scale topographic survey, realize automatic processing of low-altitude photogrammetric data using the fewest ground control points, and broaden the application area of low-altitude photogrammetry in national economical and defense construction.

解析空中三角测量是准确影像定向和高精度目标定位的最好方法。由于低空航摄影像变形大、倾角大、重叠度不规范，加之像片数量多、航线弯曲度大等特点，若采用经典空中三角测量方法，则需要大量的地面控制点，而且自动化程度低。本项目在多年研究GPS辅助空中三角测量的基础上，率先提出将北斗动态定位技术引入低空摄影测量中，通过对北斗精密单点定位模型、整周相位模糊度在航解算、多传感器集成系统误差补偿、非常规航摄影像自动量测、北斗辅助光束法平差严密数学模型及区域网优化设计等的深入研究，研制实验软件原型系统，为提高北斗/GPS系统定位摄站的精度、进而以北斗替代GPS奠定理论基础和提供实用方案，并将加密区四角地面控制点下的低空摄影测量加密提高到像素级精度，充分发挥低空摄影测量在应急与大比例尺地形测绘中的优势，实现最少量地面控制点的低空摄影测量自动数据处理，拓宽北斗卫星导航系统在国民经济和国防建设中的应用领域

本项目在多年研究GPS辅助空中三角测量的基础上，率先将北斗动态定位技术引入低空摄影测量加密中，通过对北斗（BDS）动态精密单点定位模型、基于图论和PCA-SIFT匹配方法的低空航摄影像自动量测方法、BDS辅助光束法区域网平差严密数学模型及区域网优化设计方案等的深入研究，提出了一系列专门算法并开发了实验软件Imagination系统，为提高BDS/GPS系统定位摄站的精度、进而以BDS替代GPS进行空中三角测量奠定了理论基础，同时也提供了实用技术支撑。依据本项目的研究成果，课题组在嵩山航空试验场实施了国际上首次北斗辅助无人机航摄影像的空中三角测量实验，获得了实地上加密点的平面中误差小于±0.17 m、高程中误差小于±0.25 m的最好加密精度，完全满足了丘陵地1:500比例尺地形测图的摄影测量加密精度要求，充分发挥了低空摄影测量在应急与大比例尺地形测绘中的潜力，自主实现了最少量地面控制点的低空摄影测量自动数据处理，拓宽了北斗卫星导航系统在国民经济和国防建设中的应用领域。以本项目研究成果为基础，课题组还延伸研究了采用无人机搭载可见光数码相机进行电力线自动巡检的方法，能够自动发现和准确定位电力线走廊内的障碍物，为超高压架空输电线路的安全运营提供了一种新的障碍物定量巡检手段。经过课题组成员的四年刻苦攻关，本项目共发表了14篇研究论文（其中7/11篇被SCI/EI收录），培养了4名博士、7名硕士。总体说来，项目忠实执行了立项研究计划，超额完成了既定的各项研究任务，取得了预期的研究成果。