A Dynamic Approach for Automated Detection and Cracks Using Machine Learning

Abstract

Due to severe deterioration of concrete surface infrastructure assets require frequent inspection and repair. Furthermore, unavoidable circumstances such as road accidents may occur because of defected infrastructure. Therefore, a proper civil infrastructure inspection system is essential to avoid unwanted circumstances as well as prevent traffic disruption. Manual inspection has been employed for a long time using heavy and large equipment by civil engineers to assess the structural defects. The time- consuming and labor-intensive nature of this type of inspection system causes traffic disruption. Furthermore, the manual assessment procedure is perilous for humans in inaccessible regions of civil infrastructures such as under bridge decks and underwater beams. On the contrary, an autonomous civil infrastructure inspection system monitors structural health continuously with the least human intervention. Such an independent automated framework can catch information for surface-level visual examination and deformity ID of common foundations.