Current inspection practices are often conducted manually, which is both costly and time-consuming. Thus, Unmanned Aerial Systems (UAS) present a valuable support tool for structural inspectors by reducing these inspection times and associated downtime. However, the usage of UAS in visual inspection missions also comes with significant disadvantages in terms of usable sensors and their precision.
This paper presents the concept and the prototype of a UAS optimized for vision-based inspection missions along with a novel mounting system for different sensors and cameras such as RGB- thermal or hyperspectral cameras. To measure the precision of the gathered images the spatial resolution of a mounted camera is analyzed by calculating the Spatial Frequency Response (SFR) using a specific camera test chart and an algorithm based on the slanted edge method. Also the influence of different flight maneuvers, such as hovering and flybys, as well as vibration measurements of the UAS are taken into account. Finally, spatial resolutions on the real test chart are calculated to determine the size of damages on infrastructures, that can be detected by the deployed system. Depending on lighting, camera exposure times, image angles, working distance and the considered flight maneuver the image taken by the UAS is evaluated and compared to a reference image recorded with a tripod. This study underscores the potential of UAS usage in inspection tasks, while providing critical insights into optimizing sensor performance for high-quality data acquisition.