Comparative analysis of twelve transfer learning models for the prediction and crack detection in concrete dams,based on borehole images  

作　　者：Umer Sadiq KHAN Muhammad ISHFAQUE Saif Ur Rehman KHAN Fang Xu Lerui CHEN Yi LEI 

机构地区：[1]School of Computer and Information Science,Hubei Engineering University,Xiaogan 432000,China [2]Institute for AI Industrial Technology Research,Hubei Engineering University,Xiaogan 432000,China [3]College of Water Conservancy and Hydropower Engineering,Hohai University,Nanjing 210098,China [4]School of Computer Science and Engineering,Central South University,Changsha 410083,China [5]College of Aviation,Zhongyuan University of Technology,Zhengzhou 451191,China [6]School of Civil Engineering,Central South University,Changsha 410083,China

出　　处：《Frontiers of Structural and Civil Engineering》2024年第10期1507-1523,共17页结构与土木工程前沿（英文版）

基　　金：supported by the National Natural Science Foundation of China(Grant Nos.61972136,41874148,and 42174178);the Natural Science and Foundation of Hubei Province(No.2020CFB497);the Hubei Provincial Department of Education Outstanding Youth Scientific Innovation Team Support Foundation(Nos.T201410 and T2020017);the Natural Science Foundation of Education Department of Hubei Province(No.B2020149);the Science and Technology Research Project of the Education Department of Hubei Province(No.Q20222704);Natural Science Foundation of Xiaogan City(Nos.XGKJ2022010095 and XGKJ2022010094);The funding is a foreign expert project of Henan Province(No.HNGD2023027).

摘　　要：Disaster-resilient dams require accurate crack detection,but machine learning methods cannot capture dam structural reaction temporal patterns and dependencies.This research uses deep learning,convolutional neural networks,and transfer learning to improve dam crack detection.Twelve deep-learning models are trained on 192 crack images.This research aims to provide up-to-date detecting techniques to solve dam crack problems.The finding shows that the EfficientNetB0 model performed better than others in classifying borehole concrete crack surface tiles and normal(undamaged)surface tiles with 91%accuracy.The study’s pre-trained designs help to identify and to determine the specific locations of cracks.

关 键 词：concrete dam borehole closed-circuit television deep learning models crack detection water resources management management 

分 类 号：O34[理学—固体力学]