a Light-Weight Laser Scanner for Uav Applications

a Light-Weight Laser Scanner for Uav Applications
复制标题

适用于无人机应用的轻型激光扫描仪

DOI:
10.5194/isprs-archives-xli-b1-711-2016
复制
发表时间:
2016
期刊:
ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
影响因子:
--
通讯作者:
Fernanda Magri Torres
Fernanda Magri Torres
中科院分区:
--
文献类型:
--
作者:
A. Tommaselli;Fernanda Magri Torres

文献摘要

被引文献

相似文献

无人机(UAV)由于其灵活性和良好的成本效益比,已被公认为是一种地理空间数据获取工具。无人机机载激光扫描设备的实际使用也随着新的实验和商业系统的发展而发展。本文描述了一种轻型激光扫描系统,由IbeoLux扫描仪、Novatel的惯性导航系统Span-IGM-S1、Raspberry PI便携式计算机组成,该便携式计算机记录来自两个系统和一架Octopter UAV的数据。使用地面控制点(GCP)作为参考,对该轻型系统的性能进行了准确性和点密度评估。用携带设备的无人驾驶飞行器进行了两次飞行。在第一次试验中,飞行高度为100米,在停车场上方有六个条带。第二次试验是在一个城市公园进行的,一些建筑物和人造目标作为参考地面控制点。在该实验中,选择70米的飞行高度来提高目标响应。根据实地测量的控制点坐标评估准确度。结果表明,该原型的垂直精度约为30厘米,这对于森林应用是可以接受的,但这种精度可以通过进一步改进直接地理参考和系统校准来提高。
Unmanned Aerial Vehicles (UAV) have been recognized as a tool for geospatial data acquisition due to their flexibility and favourable cost benefit ratio. The practical use of laser scanning devices on-board UAVs is also developing with new experimental and commercial systems. This paper describes a light-weight laser scanning system composed of an IbeoLux scanner, an Inertial Navigation System Span-IGM-S1, from Novatel, a Raspberry PI portable computer, which records data from both systems and an octopter UAV. The performance of this light-weight system was assessed both for accuracy and with respect to point density, using Ground Control Points (GCP) as reference. Two flights were performed with the UAV octopter carrying the equipment. In the first trial, the flight height was 100 m with six strips over a parking area. The second trial was carried out over an urban park with some buildings and artificial targets serving as reference Ground Control Points. In this experiment a flight height of 70 m was chosen to improve target response. Accuracy was assessed based on control points the coordinates of which were measured in the field. Results showed that vertical accuracy with this prototype is around 30 cm, which is acceptable for forest applications but this accuracy can be improved using further refinements in direct georeferencing and in the system calibration.