Quantitative Evaluation of Point Cloud Deformations on Glass Surfaces Caused by Unmanned Aerial Vehicle Light Detection and Ranging Incidence Angle

초록

Unmanned aerial vehicle (UAV) Light Detection and Ranging (LiDAR) is highly effective for the precise three-dimensional data collection of urban infrastructure and buildings, making it applicable in a variety of environments. However, the deformation-caused by the high transmittance and reflectance of glass, which vary with the LiDAR incidence angle-limits the accuracy and usability of the point cloud. In this study, we quantitatively evaluated the geometric deformation and intensity variation of glass surfaces with respect to the incidence angle of UAV LiDAR. For this research, point clouds of glass surfaces and control wall surfaces were collected at 10 degrees incidence angle intervals from 0 degrees to 70 degrees, with the laser pulses being approximately perpendicular to the glass surface at 0 degrees. Subsequently, the deformation of the glass surfaces was evaluated on the basis of calculated measurement errors and the extracted intensity for each incidence angle. The evaluation results revealed that changes in the incidence angle of UAV LiDAR affect the geometric deformation and intensity of the glass surface point clouds. Notably, the geometric deformation of the point clouds was minimized when the laser pulses were almost perpendicular to the glass surface, although the intensity spectrum broadened. Additionally, as the incidence angle increased, geometric deformation intensified, and the intensity approached zero. On the basis of these phenomena, we propose guidelines for setting the incidence angle of LiDAR to suit scanning purposes.

키워드