«Previous Article|Table of Contents|Next Article»

¡¶½¨Öþ¿ÆѧÓ빤³Ìѧ±¨¡·[ISSN:1673-2049/CN:61-1442/TU]

Issue:

2026Äê01ÆÚ

Page:

67-74

Research Field:

ÖÇÄܼì²âÓ뽨Ôì¼¼ÊõרÀ¸

Publishing date:

Info

Title:

Intelligent identification method of apparent defects in underwater concrete structures based on deep learning

Author(s):

CHEN Dejin;  ZHANG Donglin;  YE Xijun;  ZHENG Wenzhi;  ZHOU Junyong

School of Civil Engineering and Transportation, Guangzhou University, Guangzhou 510006, Guangdong, China

Keywords:

underwater concrete;  YOLOv9;  image fusion;  twoª²stage recognition

PACS:

TU375

DOI:

10.19815/j.jace.2024.07024

Abstract:

In underwater environment, the scattering and absorption of water media and the influence of water flow, it is easy to cause blurring, atomization, distortion and low resolution of underwater images, which subsequently reduces the accuracy of automated structural defect identification based on deep learning. Through twoª²stage processing, a model integrating image fusion algorithm and target recognition algorithm was established. In stage 1, the image fusion algorithm was used to enhance the detectability of defects. The algorithm achieved feature image weight fusion using an improved CycleGAN and a multiª²scale retinex algorithm with color protection (MSRCP). The improvements to CycleGAN included three parts, which were optimizing the cycle consistency loss, replacing the activation function, and incorporating a new attention mechanism. These enhancements could effectively enhance the image feature extraction capability, facilitate a more efficient training process, and reduce model weights. Stage 2 employed the YOLOv9 object recognition algorithm to achieve rapid and accurate identification and classification of underwater structural surface defect types in complex water environment. The results show that by enhancing underwater structural surface defect images, the mAP0.5 identification rate of the proposed method reaches 74.3%, which is 11.6% higher than that of the untreated case. The recognition accuracy of the proposed method is 74.5 %, 58.5 % and 90.0 % for the recognition of three classic apparent diseases of ¡°concrete spall¡±, ¡°exposed rebar¡±, and ¡°concrete cracking¡±, and compared with the untreated case, it is increased by 1.9 %, 5.6 % and 26.1 %, respectively.

References:

£Û 1 £Ý KOCH C, GEORGIEVA K, KASIREDDY V, et al. A review on computer vision based defect detection and condition assessment of concrete and asphalt civil infrastructure £Û J £Ý . Advanced Engineering Informatics, 2015, 29(2): 196 - 210.

£Û 2 £Ý AHLBORN T M, HARRIS D K, BROOKS C, et al. Remote sensing technologies for detectig bridge deterioration and condition assessment £Û EB/OL £Ý . (2021 ?- 07 ?- 16) £Û 2025 - 07 - 14 £Ý . https://scispace.com/papers/remote - sensing technologies - for - detecting - bridge - 17n6 drq575.

£Û 3 £Ý WU C F, SUN K K, XU Y M, et al. Concrete crack detection method based on optical fiber sensing network and microbending principle £Û J £Ý . Safety Science, 2019, 117: 299 - 304.

£Û 4 £Ý JUNG J Y, YOON H J, CHO H W. Research of remote inspection method for river bridge using sonar and visual system £Û J £Ý . Journal of the Korea Academia - industrial Cooperation Society, 2017, 18(5): 330 - 335.

£Û 5 £Ý HOU S T, JIAO D, DONG B, et al. Underwater inspection of bridge substructures using sonar and deep convolutional network £Û J £Ý . Advanced Engineering Informatics, 2022, 52: 101545.

£Û 6 £Ý ?GHANI A S A, ISA N A M. Homomorphic filtering with image fusion for enhancement of details and homogeneous contrast of underwater image £Û J £Ý .Indian Journal of Geo - marine Sciences, 2015, 44(12): 1904 - 1919.

£Û 7 £Ý LAND E H, MCCANN J J. Lightness and retinex theory £Û J £Ý . Journal of the Optical Society of America, 1971, 61(1): 1 - 11.

£Û 8 £Ý PAN X X, LI C L, PAN Z G, et al. Low ?- light image enhancement method based on retinex theory by improving illumination map £Û J £Ý . Applied Sciences, 2022, 12(10): 5257.

£Û 9 £Ý Ñîîå , ÐíÇå·ç , Íõ׿ÁÕ . »ùÓÚ¾í»ýÉñ¾­ÍøÂçµÄ½á¹¹ËðÉËʶ±ðÑо¿½øÕ¹£Û J £Ý . ½¨Öþ¿ÆѧÓ빤³Ìѧ±¨ ,2022, ª© 39(4) ªª :38 - 57.

ª¤ YANG Shuo, XU Qingfeng, WANG Zhuolin. Research progress on structural damage detection based on convolutional neural networks £Û J £Ý . Journal of Architecture and Civil Engineering, 2022, 39(4): 38 - 57.

£Û 10 £Ý ¶¡ Íþ , Âíº¥²¨ , Êæ½­Åô , µÈ . »ùÓڲвîÍøÂçµÄ»ìÄýÍÁ½á¹¹²¡º¦·ÖÀàʶ±ðÑо¿£Û J £Ý . ½¨Öþ¿ÆѧÓ빤³Ìѧ±¨ ,2022,39(4):127 - 136.

ª¤ DING Wei, MA Haibo, SHU Jiangpeng, et al. Research on classification and recognition of concrete structure diseases based on residual network £Û J £Ý . Journal of Architecture and Civil Engineering, 2022, 39(4): 127 - 136.

£Û 11 £Ý WANG Y, ZHANG J, CAO Y, et al. A deep CNN method for underwater image enhancement £Û C £Ý //IEEE. 2017 IEEE International Conference on Image Processing (ICIP). Beijing: IEEE, 2017: 1382 - 1386.

£Û 12 £Ý GUO T X, WEI Y H, SHAOH, et al. Research on underwater target detection method based on improved MSRCP and YOLOv3 £Û C £Ý //IEEE. 2021 IEEE International Conference on Mechatronics and Automation (ICMA). Takamatsu: IEEE, 2021: 1158 - 1163.

£Û 13 £Ý ÏÄ ¼á , ÖÜÀû¾ý , ÕŠΰ . »ùÓÚǨÒÆѧϰÓë VGG16 Éî¶ÈÉñ¾­ÍøÂçµÄ½¨ÖþÎïÁÑ·ì¼ì²â·½·¨£Û J £Ý . ¸£½¨½¨Éè¿Æ¼¼ ,2022(1):19 - 22,60.

ª¤ XIA Jian, ZHOU Lijun, ZHANG Wei, et al. Building crack detection method based on transfer learning and VGG16 deep neural network £Û J £Ý . Fujian Construction Science & Technology, 2022(1): 19 - 22, 60.

£Û 14 £ÝÖÜ ÖÐ , ãÆÁú±ö , ÕÅ¿¡½Ü , µÈ . »ùÓÚ YOLOX - G Ëã·¨µÄËíµÀÁÑ·ìʵʱ¼ì²â£Û J £Ý . ÌúµÀ¿ÆѧÓ빤³Ìѧ±¨ ,2023,20(7):2751 - 2762.

ª¤ ZHOU Zhong, YAN Longbin, ZHANG Junjie, ª© et al ªª . Real- time detection of tunnel cracks based on YOLOX - G algorithm £Û J £Ý . Journal of Railway Science and Engineering, 2023, 20(7): 2751 - 2762.

£Û 15 £Ý ZHU J Y, PARK T, ISOLA P, et al. Unpaired image - to - image translation using cycle - consistent adversarial networks £Û C £Ý //IEEE. 2017 IEEE International Conference on Computer Vision (ICCV). Venice: IEEE, 2017: 2242 - 2251.

£Û 16 £Ý LUCAS A, LOPEZ - TAPIA S, MOLINA R, et al. Generative adversarial networks and perceptual losses for video super- resolution £Û J £Ý . IEEE Transactions on Image Processing, 2019, 28(7): 3312 - 3327.

£Û 17 £Ý LI M, LI M S, LIU B L, et al. Spatio - temporal traffic flow prediction based on coordinated attention £Û J £Ý . Sustainability, 2022, 14(12): 7394.

£Û 18 £Ý MA R, WANG J, ZHAO W, et al. Identification of maize seed varieties using MobileNetV2 with improved attention mechanism CBAM £Û J £Ý . Agriculture, 2023, 13(1): 11.

£Û 19 £Ý SAHA R, KIM K D. Add weighted algorithm based on the PICA and RBF neural network for image fusion £Û C £Ý //IEEE. 2017 International Conference on ª© Electrical ªª , Computer and Communication Engineering (ECCE). Cox ¬ð s Bazar: IEEE, 2017: 784 - 787.

£Û 20 £Ý YANG M, SOWMYA A. An underwater color image quality evaluation metric £Û J £Ý . IEEE Transactions on Image Processing, 2015, 24(12): 6062 - 6071.

£Û 21 £Ý PANETTA K, GAO C, AGAIAN S. Human - visual - system - inspired underwater image quality measures £Û J £Ý . IEEE Journal of Oceanic Engineering, 2016, ª© 41(3) ªª : 541 - 551.

Memo

Memo:

-

Common functions

Last Update: 2026-01-20