Search & filter

Search for a publication

Search & filter

Year

Author

City

Country

New Developmentof AI crack detection system for alertagainst tunnel face collapse

Developmentof AI crack detection system for alertagainst tunnel face collapse

Developmentof_AI_Crack_Detection

Shunsuke U. / Dohta A. / Ryo H. / Hiroaki I. / Yasuji M.

Since the collapse time after cracks occur on a shotcrete surface of a tunnel face is very short. The evacuation time from the collapse of the tunnel face is limited. Therefore, to prevent disasters, AI image analysis can be effectively used to quickly recognize cracks and raise alarm. However, it is difficult to detect only on-site cracks. Therefore, we examined an AI model that can avoid noise and quickly recognize cracks. Deep leaning by segmentation method was adopted to build the AI model. Building an AI model requires a wealth of training data. The cracks drawn by the image editing software were used as training data. We have developed a system that can detect cracks on shotcrete surfaces instantly using AI image analysis.

More details

10,00 €

Year 2022
City Copenhagen
Country Denmark
ISBN 978-2-9701436-7-3