机载激光地形测绘系统的误差来源分析及消除方法

发布时间：2018-03-03 02:15

本文选题：IMU/DGPS　切入点：机载激光系统　出处：《华南理工大学》2013年硕士论文　论文类型：学位论文

【摘要】：由于传统航空摄影测量在诸多方面存在局限性，如作业时间限制，一般只能在白天进行；后期数据处理不够自动化，需要人工参与。自20世纪80年代以来，，欧美和日本相继出现各种机载激光地形测绘系统，激光系统构建容易、成本低，地形扫描快速，环境适应力强，利用激光测绘系统可以生成较为精确的数字地面模型和数字高程模型，目前已被广泛应用于路线勘探、海岸线测绘、森林资源分布测绘、数字城市建设、LBS位置服务等诸多领域。本文首先介绍了机载激光系统组成，着重于激光系统生成点云数据的理论及软件实现，包括数据通信链路协议、不同传感器数据采集的同步处理、对姿态和位置的扩展卡尔曼滤波、三维点云生成处理步骤、Delaunay三角化算法、OpenGL的三维显示与选择反馈等，即给出了一个典型的基于小型无人直升机的激光测绘系统的实现方案。由于机载激光系统上面涉及传感器众多，软件处理过程复杂，不免在结果集即三维点云中的X、Y、Z三轴方向会产生误差，导致地形倾斜、周期性地扭曲等，因此本文分析了各个硬件与软件处理环节对三维点云的影响特征，并进行了数值模拟计算，方便指导我们在实践过程中对误差的判断与分析，最后给出了部分系统误差的检校与消除方法，包括姿态角校正、安置角恢复、平差处理等，误差包括随机误差和系统误差，通过对误差的检校，可以消除大部分的系统误差，提升点云精度。
[Abstract]:Because of the limitations of traditional aerial photogrammetry in many aspects, such as operating time limit, it can only be carried out during the day, and the later data processing is not automatic enough and requires manual participation. Since 1980s, A variety of airborne laser topographic mapping systems have emerged in Europe, America and Japan one after another. The laser system is easy to build, low in cost, rapid in topographic scanning, and strong in environmental adaptability. Laser mapping system can generate more accurate digital ground model and digital elevation model, which has been widely used in many fields such as route exploration, shoreline mapping, forest resource distribution mapping, digital city construction and LBS location service. This paper first introduces the composition of airborne laser system, focusing on the theory and software implementation of point cloud data generation in laser system, including data communication link protocol, synchronous processing of data acquisition from different sensors. The extended Kalman filter for attitude and position, the three-dimension display and selection feedback for Delaunay triangulation algorithm, a typical laser mapping system based on small unmanned helicopter is presented. Because there are many sensors involved in the airborne laser system and the software processing process is complex, it can't avoid the error in the three axis direction of the result set, that is, the three dimensional point cloud, which leads to the terrain tilt, periodic distortion and so on. Therefore, this paper analyzes the influence characteristics of each hardware and software processing link on 3D point cloud, and carries out numerical simulation calculation, which is convenient to guide us to judge and analyze the error in the process of practice. Finally, the calibration and elimination methods of some system errors are given, including attitude angle correction, fixing angle recovery, adjustment processing and so on. The errors include random error and system error. Most of the system errors can be eliminated by checking the errors. Improve point cloud accuracy.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P231;V243

【参考文献】