Crack Detection and Width Measurement for Concrete Building using Mobile Devices

Crack detection and width measurement are critical for the long-term safe operation of concrete structures. In Japan, cracks with a width of 0.3 mm or larger are set as inspection standards because water penetration through cracks in reinforced concrete can cause corrosion of the reinforcing steel. Current inspection methods are typically manual which involve time-consuming and labor-intensive work, and there is a shortage of manpower. Additionally, measurement errors can occur depending on the measurer’s experience and the inspection environment. In previous research on the automation of crack detection and width measurement in millimeters, it was necessary to use special equipment, place markers on the measurement surface, or take photographs with the measurement surface and camera axis facing each other vertically. The purpose of this study is to make the inspection easier by using a mobile device with a built-in LiDAR and RGB camera. In this study, we proposed a method for clarifying the angle and position of the measurement plane and camera using point cloud and camera self-position estimation, and for complementing the points corresponding to the image pixels. Consequently, it is possible to measure crack width even when the surface and camera axis were tilted. Furthermore, since all the necessary data can be obtained using only mobile devices, no special equipment or wall markers are required. In order to demonstrate the effectiveness, results of proposed method were compared with the manual measurement values. As a result, the maximum error was 0.148mm, and the standard deviation decreased at all 15 measurement points. According to these results, the flexibility of shooting camera position is increased. This achievement is important in terms of simplifying the inspection of concrete structures and reducing errors.