Assessment of the feasibility of detecting concrete cracks in images acquired by unmanned aerial vehicles

摘要

An 8-rotor unmanned aerial vehicle is used as a working platform. Its motion characteristics in a hovering state are obtained using a non-contact measurement instrument, which, along with the modulation transfer function of its airborne images, indicates the reliability of the airborne images of unmanned aerial vehicles in a hovering state. By installing a laser range finder on the cradle synchronized with the camera shutter to measure the object distance, the pixel resolution of the object distance is obtained. The airborne images are then processed using the MAMAS image processing toolbox, from which the pixels of concrete cracks are extracted. Compared to a static image and direct manual measurements, the airborne image of the unmanned aerial vehicle has higher precision, indicating its wide potential applications as an alternative of the conventional inspection methods of bridge inspection vehicle and working platforms.