A post-processing algorithm for automatic recognition of tunnel crack diseases based on segmentation masks

doi:10.11947/j.AGCS.2024.20240088

Abstract

Abstract:

With the construction of transportation networks, the number of completed tunnels and the increasing service life of tunnels have brought great challenges to the safe operation of tunnels. Rapid detection of tunnel lining cracks and accurate extraction of crack length and width characteristics is an important guarantee for achieving efficient maintenance and safe operation of tunnel. This article proposes an efficient and accurate post-processing algorithm for tunnel crack diseases, based on the prediction segmentation mask of DeepLabV3+ semantic segmentation model. The connected domain discrimination refinement algorithm and endpoint clustering instance differentiation algorithm are used to process the mask fracture situation, achieving accurate extraction of tunnel crack skeleton and instance differentiation. Finally, the length calculation and grayscale difference value width classification algorithm are used to calculate the crack length and width characteristics. The accuracy of length and width calculation is 92.2% and 86.3%, respectively.

Key words: crack disease, semantic segmentation, intelligent computing, DeepLabV3+

CLC Number:

P258

Bo HU, Hanxin CHEN, Song REN, Yinghao QU, Qingyi LIU, Xinyue TU, Datao WANG. A post-processing algorithm for automatic recognition of tunnel crack diseases based on segmentation masks[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(9): 1715-1724.

Figures/Tables 16

Tab.1

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

Tab.2

Tab.3

Tab.4

Fig.10

Fig.11

Fig.12

References 25

[1]	刘海京, 夏才初, 朱合华, 等. 隧道病害研究现状与进展[J]. 地下空间与工程学报, 2007, 3(5):947-953.
	LIU Haijing, XIA Caichu, ZHU Hehua, et al. Studies on tunnel damage[J]. Chinese Journal of Underground Space and Engineering, 2007, 3(5):947-953.
[2]	吴江滨, 张顶立, 王梦恕. 铁路运营隧道病害现状及检测评估[J]. 中国安全科学学报, 2003, 13(6):49-52.
	WU Jiangbin, ZHANG Dingli, WANG Mengshu. Current damage situation of railway operation tunnels and their inspection and evaluation[J]. China Safety Science Journal, 2003, 13(6):49-52.
[3]	SHEN Yueqian, WANG Jinguo, WANG Jinhu, et al. Methodology for extraction of tunnel cross-sections using dense point cloud data[J]. Journal of Geodesy and Geoinformation Science, 2021, 4(2):56-71.
[4]	王耀东, 朱力强, 余祖俊, 等. 基于样本自动标注的隧道裂缝病害智能识别[J]. 西南交通大学学报, 2023, 58(5):1001-1008.
	WANG Yaodong, ZHU Liqiang, YU Zujun, et al. Intelligent tunnel crack recognition based on automatic sample labeling[J]. Journal of Southwest Jiaotong University, 2023, 58(5):1001-1008.
[5]	张振海, 尹晓珍, 王阳萍, 等. 基于特征分析的图像式地铁隧道裂缝检测方法研究[J]. 铁道科学与工程学报, 2019, 16(11):2791-2800.
	ZHANG Zhenhai, YIN Xiaozhen, WANG Yangping, et al. Research on image-based crack detection method for subway tunnel based on feature analysis[J]. Journal of Railway Science and Engineering, 2019, 16(11):2791-2800.
[6]	王耀东, 余祖俊, 白彪, 等. 基于图像处理的地铁隧道裂缝识别算法研究[J]. 仪器仪表学报, 2014, 35(7):1489-1496.
	WANG Yaodong, YU Zujun, BAI Biao, et al. Research on image processing based subway tunnel crack identification algorithm[J]. Chinese Journal of Scientific Instrument, 2014, 35(7):1489-1496.
[7]	朱苦竹, 庄宁. 公路隧道衬砌裂缝成因分析及其治理[J]. 桂林工学院学报, 2009, 29(3):354-359.
	ZHU Kuzhu, ZHUANG Ning. Analysis and solution to lining crack in one tunnel[J]. Journal of Guilin University of Technology, 2009, 29(3):354-359.
[8]	何国华, 刘新根, 陈莹莹, 等. 基于数字图像的隧道表观病害识别方法研究[J]. 重庆交通大学学报(自然科学版), 2019, 38(3):21-26.
	HE Guohua, LIU Xingen, CHEN Yingying, et al. Apparent tunnel diseases identification based on digital images[J]. Journal of Chongqing Jiaotong University (Natural Science), 2019, 38(3):21-26.
[9]	王平让, 黄宏伟, 薛亚东. 基于图像局部网格特征的隧道衬砌裂缝自动识别[J]. 岩石力学与工程学报, 2012, 31(5):991-999.
	WANG Pingrang, HUANG Hongwei, XUE Yadong. Automatic recognition of cracks in tunnel lining based on characteristics of local grids in images[J]. Chinese Journal of Rock Mechanics and Engineering, 2012, 31(5):991-999.
[10]	雷斯达, 曹鸿猷, 康俊涛. 基于深度学习的复杂场景下混凝土表面裂缝识别研究[J]. 公路交通科技, 2020, 37(12):80-88.
	LEI Sida, CAO Hongyou, KANG Juntao. Study on concrete surface crack recognition in complex scenario based on deep learning[J]. Journal of Highway and Transportation Research and Development, 2020, 37(12):80-88.
[11]	李健超, 张翠兵, 柴雪松, 等. 基于图像识别技术的隧道衬砌裂缝检测系统研究[J]. 铁道建筑, 2018, 58(1):20-24.
	LI Jianchao, ZHANG Cuibing, CHAI Xuesong, et al. Research on crack detection system of tunnel lining based on image recognition technology[J]. Railway Engineering, 2018, 58(1):20-24.
[12]	周中, 张俊杰, 鲁四平. 基于改进YOLOv4的隧道衬砌裂缝检测算法[J]. 铁道学报, 2023, 45(10):162-170.
	ZHOU Zhong, ZHANG Junjie, LU Siping. Tunnel lining crack detection algorithm based on improved YOLOv4[J]. Journal of the China Railway Society, 2023, 45(10):162-170.
[13]	薛亚东, 李宜城. 基于深度学习的盾构隧道衬砌病害识别方法[J]. 湖南大学学报(自然科学版), 2018, 45(3):100-109.
	XUE Yadong, LI Yicheng. A method of disease recognition for shield tunnel lining based on deep learning[J]. Journal of Hunan University (Natural Sciences), 2018, 45(3):100-109.
[14]	吴刚, 罗炜, 王小龙, 等. 基于深度学习的盾构隧道衬砌表观病害检测模型研究[J]. 现代隧道技术, 2023, 60(4):67-75.
	WU Gang, LUO Wei, WANG Xiaolong, et al. Study on a deep learning-based model for detecting apparent defects in shield tunnel lining[J]. Modern Tunnelling Technology, 2023, 60(4):67-75.
[15]	WANG Hanxiang, LI Yanfen, DANG L M, et al. Pixel-level tunnel crack segmentation using a weakly supervised annotation approach[J]. Computers in Industry, 2021, 133:103545.
[16]	ZHAO Shuai, ZHANG Dongming, XUE Yadong, et al. A deep learning-based approach for refined crack evaluation from shield tunnel lining images[J]. Automation in Construction, 2021, 132:103934.
[17]	朱磊, 李东彪, 闫星志, 等. 基于改进Mask R-CNN深度学习算法的隧道裂缝智能检测方法[J]. 图学学报, 2023, 44(1):177-183.
	ZHU Lei, LI Dongbiao, YAN Xingzhi, et al. Intelligent detection method of tunnel cracks based on improved Mask R-CNN deep learning algorithm[J]. Journal of Graphics, 2023, 44(1):177-183.
[18]	HUANG Hongwei, ZHAO Shuai, ZHANG Dongming, et al. Deep learning-based instance segmentation of cracks from shield tunnel lining images[J]. Structure and Infrastructure Engineering, 2022, 18(2):183-196.
[19]	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of 2017 IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2980-2988.
[20]	MILLETARI F, NAVAB N, AHMADI S A. V-net: fully convolutional neural networks for volumetric medical image segmentation[C]//Proceedings of 2016 International Conference on 3D Vision. Stanford: IEEE, 2016: 565-571.
[21]	拉维尚卡·奇特亚拉, 斯林德维·普迪佩迪. Python图像处理与采集[M]. 北京: 清华大学出版社, 2023: 22.
	RAVISHANKA Chityala, SLINDEVI Pudipedi. Python image processing and acquisition[M]. Beijing: Tsinghua University Press, 2023: 22.
[22]	耿楠. 数字图像处理[M]. 4版. 西安: 西安电子科技大学出版社, 2022: 272.
	GENG Nan. Digital image processing[M]. 4th ed. Xi'an: Xidian University Press, 2022: 272.
[23]	李淳罡. 第8章：形态学操作[EB/OL]. (2022-01-23)[2023-03-05]. https://blog.csdn.net/weixin_57440207/article/details/122647000?spm=1001.2014.3001.5502.
	LI Chungang. Chapter 8: morphological operations[EB/OL]. (2022-01-23)[2023-03-05]. https://blog.csdn.net/weixin_57440207/article/details/122647000?spm=1001.2014.3001.5502.
[24]	YUAN . 自动驾驶路径规划——A*算法[EB/OL]. [2023-03-05]. https://blog.csdn.net/sinat_52032317/article/details/127077625.
	YUAN . Autonomous driving path planning--A* algorithm[EB/OL]. [2023-03-05]. https://blog.csdn.net/sinat_52032317/article/details/127077625.
[25]	KANG D, BENIPAL S S, GOPAL D L, et al. Hybrid pixel-level concrete crack segmentation and quantification across complex backgrounds using deep learning[J]. Automation in Construction, 2020, 118:103291.

输入尺寸	操作结构	扩展因子	膨胀率	输出矩阵深度
1024×1024×3	convd2d	—	—	32
512×512×32	inverted bottleneck	1	1	16
512×512×16	inverted bottleneck×2	6	1	24
256×256×24	inverted bottleneck×3	6	1	32
128×128×32	inverted bottleneck×4	6	2	64
128×128×64	inverted bottleneck×3	6	2	96
128×128×96	inverted bottleneck×3	6	4	160
128×128×160	inverted bottleneck	6	4	320

g₁取值	裂缝宽度类别
g₁<Gray1	0.2 mm≤W<0.6 mm
Gray1≤g₁≤Gray2	0.6 mm≤W≤1 mm
g₁>Gray2	W>1 mm

参数	取值
取值定位点数量N	10，裂缝骨架像素点数量小于等于200时 40，裂缝骨架像素点数量大于200时
单侧灰度取值点数量X	10.0
Gray1	11.7
Gray2	19.0

数据集	准确率	召回率	L_R/(%)	W_R/(%)	每张图处理时间/s
训练集	0.913	0.949	93.2	88.5	—
测试集	0.914	0.936	92.2	86.3	0.449