Unmanned aerial vehicle (UAV)-powered concrete crack detection based on digital image processing

Abstract

Civil infrastructure can suffer from concrete cracks due to the creep, shrinkage, diverse natural loading, aging, and corrosion of reinforcement. Concrete cracks are one of the important features when analyzing the current condition of concrete structures. The shape, length, and width information on concrete cracks are possible to conduct an initial maintenance for the structure safety. Generally, concrete cracks in structures are inspected by the inspector who evaluates the structural health by observing and measuring visual damages on the surface of concrete structures. However, the visual inspection is time consuming and costly, it is dangerous for specialists, and improper environments can cause evaluation errors. Recently, digital image processing (DIP) has been introduced as an alternative automatic crack assessment to the visual inspection. However, image processing has some issues to conduct the crack identification of large-scale infrastructure in terms of the time, cost, and safety. Recent advancements in UAV technologies have enabled low-cost, high-performance UAVs that can be adopted in diverse technologies. In this paper, a prototype of the UAV-based image processing is proposed to identify concrete cracks. From the field experiment, the calculated crack width is similar with the measured crack width by crack gauges.