| After an earthquake, roads need to beopened to traffic as soon as possible for emergency transportation. To this end, it is necessary to quickly and accurately assess the damage to road bridges and confirm the structural safety and serviceability of bridges. However, the current post-seismic inspections are first conducted by vehicle patrols, and if there are any deformities on the road surface, a hands-on inspection is conducted. This inspection method is time consuming. Therefore, this study was conducted with the aim of speeding up post-quake inspections by utilizing images taken by UAVs. When inspections are supplemented by technologies such as UAVs, especially in the case of post-seismic inspections, it is necessary to perform diagnostics regarding structural safety, so missing critical damage is unacceptable. Therefore, it was first necessary to assess what kind of imaging conditions would ensure which damage could be recognized from the images. Therefore, an experiment was conducted to clarify the relationship between shooting distance and identification accuracy for each type of earthquake damage. As a result, we showed the performance curves of the inspection index and shooting distance. For example, in a typical performance UAV, the results showed that at a shooting distance of 10 m from the object, a crack width of more than 2 mm, a deformation angle of more than 5°, and a deformation amount of more than 10 mm could be expected to be captured. |
