程浩东,李怡静,李玥康,胡强,王姣

• 1:南昌大学工程建设学院
• 2:江西省水利科学院

摘要(Abstract)：

【目的】背景复杂的混凝土结构表面裂缝连续性差、识别率低，基于深度学习的裂缝检测方法存在模型参数量大的问题。【方法】为此，结合U~2-Net框架构建了一种聚合多尺度信息的轻量级模型U~2-Net＿Aggregation,用于复杂背景下的裂缝特征学习。该模型通过增加跳跃连接，使得每个解码层均聚合该层以上所有浅层编码特征，以获得足够的特征细节，提升裂缝分割精度；利用深度可分离卷积(Depthwise Separable Convolution, DSC)对原本的残差模块(ReSidual U-blocks, RSU)进行改进，提出了新的残差模块(RSU-DSC-ECA),来降低聚合多尺度信息时带来的模型复杂度提升的问题，其中的通道注意力机制(Efficient Channel Attention, ECA)可提升模型对裂缝区域的敏感性和对复杂背景的抗干扰能力。【结果】在三组裂缝数据集上进行消融试验，改进后的模型(U~2-Net＿Aggregation)相较于U~2-Net在准确率、交并比、综合评价指标上均有优异的表现。为了验证模型对复杂背景中裂缝的识别能力，利用无人机实地采集的某混凝土结构数据进行试验，其检测效果优于FCN、SegNet、U-Net和U~2-Net。【结论】改进后的模型相比U~2-Net在召回率、交并比和综合评价指标方面分别提高了4.18%、2.97%和2.03%,可借助无人机影像快速准确地检测出裂缝，为结构裂缝检测提供一种新的方法。

关键词(KeyWords)： 混凝土结构;裂缝检测;深度学习;语义分割;U~2-Net;神经网络;混凝土

Abstract:

Keywords:

基金项目(Foundation): 江西省自然科学基金项目(20232BAB204091);; 国家自然科学基金项目(41501454);; 江西省水利厅科技项目(202123YBKT25)

作者(Author): 程浩东,李怡静,李玥康,胡强,王姣

DOI: 10.13928/j.cnki.wrahe.2024.06.013

参考文献(References)：

• [1] 王丽.混凝土坝裂缝识别的数字图像处理算法研究 [D].大连：大连理工大学，2019.WANG Li.Research on Digital Image Processing Algorithms for Crack Recognition of Concrete Dams[D].Dalian:Dalian University of Technology,2019.
• [2] DONG C Z,CATBAS F N.A review of computer vision-based structural health monitoring at local and global levels [J].Structural Health Monitoring,2021,20(2):692-743.
• [3] 徐辉，贺俊吉，方亮.基于图像显著性的大坝裂缝检测 [J].工业控制计算机，2018,31(9):19-20.XU Hui,HE Junji,FANG Liang.Dam crack detection based on image saliency[J].Industrial Control Computer,2018,31(9):19-20.
• [4] 陈伟，范新南，李敏，等.基于Gabor算子的人工蜂群算法大坝裂缝检测 [J].微处理机，2015,36(4):32-38.CHEN Wei,FAN Xinnan,LI Min,et al.Dam crack detection of artificial colony algorithm based on gabor operator[J].Microprocessors,2015,36(4):32-38.
• [5] 张大伟，许梦钊，马莉，等.水下大坝裂缝图像分割方法研究 [J].软件导刊，2016,15(9):170-172.ZHANG Dawei,XU Mengzhao,MA Li,et al.Research on image segmentation of underwater dam cracks[J].Software Guide,2016,15(9):170-172.
• [6] 张来胜，李亚萍.融合视觉处理的堤坝水下裂缝检测方法 [J].软件导刊，2022,21(6):188-193.ZHANG Laisheng,LI Yaping.Underwater crack detection method of dam with visual processing[J].Software Guide,2022,21(6):188-193.
• [7] 张小伟，包腾飞，高兴和.基于计算机视觉的混凝土坝裂缝检测方法 [J].水利水电科技进展，2021,41(5):83-88.ZHANG Xiaowei,BAO Tengfei,GAO Xinghe.Crack detection method of concrete dams based on computer vision[J].Advances in Science and Technology of Water Resources,2021,41(5):83-88.
• [8] 张小伟，包腾飞.基于局部大津阈值与区域生长的坝面细小裂缝识别分割算法 [J].水电能源科学，2022,40(2):97-100.ZHANG Xiaowei,BAO Tengfei.Identification and segmentation algorithm of small cracks on dam surface based on Local OTSU Threshold and Regional Growth[J].Water Resources and Power,2022,40(2):97-100.
• [9] 申勇，宋致军，宋方玉，等.“双碳”背景下水利工程的发展趋势与研究进展[J].水利发展研究，2023,23(6):30-36.SHEN Yong,SONG Zhijun,SONG Fangyu,etc The development trends and research progress of water conservancy engineering under the background of “Dual Carbon”[J].Water Resources Development Research,2023,23(6):30-36.
• [10] HSIEH Y A,TSAI Y J.Machine learning for crack detection:review and model performance comparison [J].Journal of Computing in Civil Engineering,2020,34(5):04020038.
• [11] YAMAGUCHI T,HASHIMOTO S.Fast crack detection method for large-size concrete surface images using percolation-based image processing [J].Machine Vision & Applications,2010,21(5):797-809.
• [12] CHUN P J,IZUMI S,YAMANE T.Automatic detection method of cracks from concrete surface imagery using two-step light gradient boosting machine [J].Computer-Aided Civil and Infrastructure Engineering,2021,36(1):61-72.
• [13] LIU Z,AZMIN S,OHASHI T,et al.Tunnel crack detection and classification system based on image processing[C]// Conference on Machine Vision Applications in Industrial Inspection X.San Jose,Ca,2002:145-152.2002:495-499.
• [14] SINHA S K,FIEGUTH P W.Automated detection of cracks in buried concrete pipe images [J].Automation in Construction,2006,15(1):58-72.
• [15] KIM H,AHN E,SHIN M,et al.Crack and noncrack classification from concrete surface images using machine learning [J].Structural Health Monitoring,2018,18(3):147592171876874.
• [16] DORAFSHAN S,THOMAS R J,MAGUIRE M.Comparison of deep convolutional neural networks and edge detectors for image-based crack detection in concrete [J].Construction and Building Materials,2018,186:1031-1045.
• [17] JU H Y,LI W,TIGHE S S,et al.CrackU-net:A novel deep convolutional neural network for pixelwise pavement crack detection [J].Structural Control & Health Monitoring,2020,27(8):e2551.1-e2551.19.
• [18] ZOU Q,ZHANG Z,LI Q Q,et al.DeepCrack:Learning hierarchical convolutional features for crack detection [J].Ieee Transactions on Image Processing,2019,28(3):1498-1512.
• [19] LIU Z Q,CAO Y W,WANG Y Z,et al.Computer vision-based concrete crack detection using U-net fully convolutional networks [J].Automation in Construction,2019,104:129-139.
• [20] YANG X C,LI H,YU Y T,et al.Automatic pixel-level crack detection and measurement using fully convolutional network [J].Computer-Aided Civil and Infrastructure Engineering,2018,33(12):1090-1109.
• [21] 陈波，张华，汪双，等.基于全卷积神经网络的坝面裂纹检测方法研究 [J].水力发电学报，2020,39(7):52-60.CHEN Bo,ZHANG Hua,WANG Shuang,et al.Study on detection method of dam surface cracks based on full convolution neural network[J].Journal of Hydroelectric Engineering,2020,39(7):52-60.
• [22] 任秋兵，李明超，沈扬，等.水工混凝土裂缝像素级形态分割与特征量化方法 [J].水力发电学报，2021,40(2):234-246.REN Qiubing,LI Mingchao,SHEN Yang,et al.Pixel-level shape segmentation and feature quantification of hydraulic concrete cracks based on digital images[J].Journal of Hydroelectric Engineering,2021,40(2):234-246.
• [23] LI S,ZHAO X,ZHOU G.Automatic pixel-level multiple damage detection of concrete structure using fully convolutional network [J].Computer-Aided Civil and Infrastructure Engineering,2019,34 (7).DOI:10.1111/mice.12433.
• [24] DUNG C V,ANH L D.Autonomous concrete crack detection using deep fully convolutional neural network [J].Automation in Construction,2018,99:52-58.
• [25] FENG C C,ZHANG H,WANG H R,et al.Automatic pixel-level crack detection on dam surface using deep convolutional network [J].Sensors,2020,20(7):1-20.
• [26] QIN X B,ZHANG Z C,HUANG C Y,et al.U2-Net:Going deeper with nested U-structure for salient object detection [J].Pattern Recognition,2020,106:1-12.
• [27] WANG L J,LU H C,WANG Y F,et al.Learning to Detect Salient Objects with Image-level Supervision[C]//IEEE.Proceedings of the 30th IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Honolulu,HI,2017:3796-3805.
• [28] ZHOU Z W,SIDDIQUEE M M R,TAJBAKHSH N,et al.U-Net plus plus :Redesigning skip connections to exploit multiscale features in image segmentation [J].IEEE Transactions on Medical Imaging,2020,39(6):1856-1867.
• [29] HUANG H M,LIN L F,TONG R F,et al.U-Net 3+:A full-scale connected unet for medical image segmentation[C] //IEEE.Proceedings of the IEEE International Conference on Acoustics,Speech,and Signal Processing,Barcelona,SPAIN,2020:1055-1059.
• [30] CHOLLET F,IEEE.Xception:Deep Learning with Depthwise Separable Convolutions[C]//IEEE.Proceedings of the 30th IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),Honolulu,HI,2017:1800-1807.
• [31] ALZOUBE M A.Efficient vision-based multi-target augmented reality in the browser [J].Multimedia Tools and Applications,2022,81(10):14303-14320.
• [32] WANG Q,WU B,ZHU P,et al.ECA-Net:Efficient Channel Attention for Deep Convolutional Neural Networks[C] //IEEE.Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),2020.
• [33] ?ZGENEL ? F,SORGU? A G.Performance comparison of pretrained convolutional neural networks on crack detection in buildings[C] //IAARC.Proceedings of the Isarc proceedings of the international symposium on automation and robotics in construction,IAARC Publications,2018:1-8.
• [34] LIU Y,YAO J,LU X,et al.DeepCrack:A deep hierarchical feature learning architecture for crack segmentation [J].Neurocomputing,2019,338:139-153.
• [35] 李怡静，程浩东，李火坤，等.基于改进U2-Net与迁移学习的无人机影像堤防裂缝检测 [J].水利水电科技进展：,2022,42(6):52-59.LI Yijing,CHENG Haodong,LI Huokun,et al.Crack detection of embankment in UAV images based on improved U2-Net and transfer learning[J].Advances in Science and Technology of Water Resources,2022,42(6):52-59.
• [36] 刘伟，邹春霞，薛慧君，等.渠道衬砌混凝土冻融风蚀耐久性能及寿命预测[J].排灌机械工程学报，2023,41(6):568-575.LIU Wei,ZOU Chunxia,XUE Huijun,et al.Durability and life prediction of freeze-thaw-wind erosion of channel lining concrete[J].Journal of Drainage and Irrigation Machinery Engineering,2023,41(6):568-575.