Laser scanning surveys conducted on complex technical infrastructure with a horizontal structure often encounter the issue of mutual obscuration of elements. This problem can be addressed by utilizing UAV surveys, provided that the necessary accuracy is maintained. Photogrammetric drones are susceptible to disturbances such as shading, lighting, and reflections, which can impact survey accuracy. Scanning drones offer more stable accuracy in such scenarios. The research aimed to determine if a scanning drone could generate point clouds of technical infrastructure with geometry accuracy comparable to terrestrial scanning. Previous studies on scanning drones have primarily focused on absolute measurement accuracy, typically within a few centimeters. However, for determining the geometric shape of installation elements in a local coordinate system, internal measurement consistency is crucial. The research demonstrated that by integrating point clouds from multiple drone flight paths effectively, consistency within a few to several millimeters can be achieved. A comparison with terrestrial laser scanning results indicated the ability to detect deformations at a level of approximately 5–10 mm.
Read full abstract