Publications / 2020 Proceedings of the 37th ISARC, Kitakyushu, Japan

Cracks Detection using Artificial Intelligence to Enhance Inspection Efficiency and Analyze the Critical Defects

Fawaz Habbal, Abdualla Alnuaimi, Mohammed Al Shamsi, Saleh Alshaibah and Thuraya Aldarmaki Thuraya.Aldarmaki
Pages 1367-1372 (2020 Proceedings of the 37th ISARC, Kitakyushu, Japan, ISBN 978-952-94-3634-7, ISSN 2413-5844)

Cracks are one of the significant criteria utilized for diagnosing the disintegration of solid structures. Normally, a structural designer with specific information would assess such structures by checking for breaks outwardly, outlining the aftereffects of the examination, and afterward getting ready to investigate information based on their discoveries. A review technique like this is not without a doubt, as it requires manual chronicle of upwards of a few hundred thousand splits individually. However, it additionally cannot precisely identify the length and state of the breaks. To handle the issue, the industry has progressively looked toward utilizing AI detection to conduct a direct based assessment. Researchers are as of now building up a mobile-based intelligence equipped with AI analytics for recognizing splits and different imperfections, which will upgrade the proficiency of the investigation process and enhance the emergency response towards users. This study aims to provide an insight on the most efficient crack detection method by comparing several techniques. It was found out that the hybrid technique was the most efficient method to detect cracks. This technique combines the positive points from both the artificial neural network (ANN) and artificial bee colony (ABC) to come up with a more proper entirely new method.

Keywords: Cracks Detection; Inspection Enhancements; AI Detection; Cracks Deep Learning; Buildings Defects