Regular inspection of tunnel ancillary facilities has become a cornerstone of sustainable tunnel operation and maintenance. Conventional manual inspections, however, are often plagued by low efficiency, elevated safety risks, and inherent subjectivity. In contrast, AI-driven diagnostic frameworks offer a transformative approach to enhancing detection precision and operational efficiency, providing a robust foundation for granular tunnel management. Given that current understanding of tunnel intelligent detection equipment and algorithms has not yet been systematized, a systematic review of recent research progress was conducted, based on a synthesis of relevant literature, from three perspectives: detection equipment, object detection algorithms, and defect detection algorithms. Regarding detection equipment, the architectural frameworks of mobile detection platforms were systematically introduced, and the advantages and limitations of two mainstream data acquisition techniques (high-resolution visible-light imaging and laser scanning) were thoroughly analyzed. Multi-sensor fusion technologies and tunnel positioning techniques were further explored. Concerning object detection algorithms, image processing techniques tailored to the special tunnel environment and data augmentation methods were summarized with typical case studies, and the performance of two-stage versus single-stage detection algorithms for tunnel ancillary facility target detection was compared, with emphasis placed on this comparison. With respect to defect detection algorithms, commonly adopted defect detection methods were reviewed, and the application of semantic segmentation and instance segmentation techniques to the refined assessment of facility defects was discussed. Finally, the core challenges confronting current intelligent equipment and technologies were summarized, including the difficulty of satisfying high-speed dynamic inspection requirements, the lack of standardized datasets, and insufficient model generalization capability. In response to these challenges, the outlook for future research was directed toward three major avenues: the development of integrated intelligent detection equipment tailored to high-speed dynamic scenarios, the exploration of generative AI for data augmentation, and the enhancement of model generalization ability and robustness under complex tunnel environmental conditions.
ZHAO Yong-zhi, ZHANG Pei-qi, ZHAO Xiao-jin, SUN Li-zhou, WANG Qing, SHI Zhen-ming.
Review of AI-based Equipment and Technologies for Tunnel Facility Inspection[J].
Science Technology and Engineering, 2026, 26(30): 12882-12901 DOI:10.12404/j.issn.1671-1815.2601199
由于光学影像成像与激光扫描技术存在各自的优点与不足,部分研究者将二者结合,以提升数据采集的准确性和完整性。其关键技术为多传感器融合,即通过算法将不同传感器数据进行融合,获得比单一数据源更为可靠、准确的数据。多传感器融合技术包含:①光学影像成像(相机),负责捕捉色彩、纹理等视觉信息,对于识别隧道设施类型和颜色缺损(如锈蚀)至关重要;②激光扫描技术,对环境光照变化不敏感,是构建高精度三维模型、测量隧道结构几何尺寸和形变的基础[31];③惯性测量单元(inertial measurement unit, IMU),以极高频率测量加速度和角速度,提供平台运动轨迹,在全球卫星导航系统(global navigation satellite system, GNSS)信号缺失的隧道环境中,IMU尤为重要;④数据融合算法,如卡尔曼滤波被广泛用于融合高频但随时间漂移的IMU数据和低频但更为精确的相机或激光扫描数据,从而补偿单传感器弱点。
基于上述多传感器融合理念,研究人员开发了多种移动式检测系统,如Menendez等[32]开发的隧道智能巡检机器人系统(robotic system with intelligent control for cave-type infrastructures and tunnels inspection, ROBO-SPECT),由车辆、起重机、机械臂、半监督计算机视觉系统及超声波传感器等组成,能实现半自动化检测裂缝宽度和深度。Ai等[33]基于CCD相机和激光扫描技术开发了地铁隧道结构检测装置,移动采集速度可达5 km/h。Nakamura等[34]研发的隧道衬砌混凝土检测车,配备机器视觉和锤击检测设备,可检测隧道衬砌表面裂缝及内部损伤,最大运行速度为10 m/min,可识别宽度≥0.5 mm的裂缝。Yasuda等[35]开发的MIMM-R隧道检测车集成了空气耦合式雷达、多台线阵CCD相机及激光扫描装置,如图5[36]所示,能对衬砌厚度、空洞、裂缝等进行检测,速度可达70 km/h。黄希等[37]集成激光扫描仪、惯性测量单元、里程传感器等多种技术,研发了轨道移动激光测量与检测系统(图6),可对铁路、地铁等不同轨道交通限界进行检测。石波等[25]将线阵相机与激光扫描技术结合,将二者坐标基准统一,并外触发使传感器采集频率与车速匹配,对轨交隧道病害进行检测定位。
对于车载式高速数据采集平台,相机与隧道壁之间的高速运动易导致图像运动模糊,使得裂缝等微小目标边界模糊不清。图像运动模糊还原方法[53-58]分类如表3所示,传统方法主要基于反卷积思想,如维纳滤波、R-L迭代还原算法(Richardson-Lucy algorithm)、基于正则化等,但这些方法通常假设模糊核是均匀的,在隧道检测场景下,常会因车辆震动、设备抖动等因素出现非均匀模糊,致使传统方法难以准确建模。深度学习方法通过端到端训练,能够学习从模糊图像到清晰图像的复杂映射,直接从模糊图像中恢复清晰纹理。特别是引入注意力机制的模型,其能够在去除运动拖影的同时,避免传统反卷积方法产生的振铃效应和伪影,显著提升采集数据的质量[59]。Ma等[60]基于隧道采集模糊图像的多尺度特征,采用尺度循环去模糊网络(scale-recurrent network for deep image deblurring, SRN-DeblurNet)方法进行去模糊化处理(图7),并基于改进的YOLOv8模型进行了隧道渗漏识别。
得益于像素级缺损轮廓划分能力,语义分割算法被广泛应用于工程缺损检测。Haltakov等[117]针对交通信号灯检测,利用语义分割算法生成候选区域,结合几何和颜色分类,实现了白天和夜晚信号灯检测。Lee等[118]基于改进的SegNet模型对裂缝进行检测,在此基础上对裂缝宽度进行了预测,借助其提出的形状敏感核函数,预测精度比传统语义分割模型精确1或2个像素。针对裂缝检测算法实现复杂、泛化能力弱、像素精度低等问题,Zhang等[119]基于UNet模型进行改进,用于自动像素级裂纹检测,研究了数据集大小和模型深度对训练时间、检测精度和速度的影响,提高了检测算法的鲁棒性和泛化能力。Chen等[120]引入轻量级语义分割模型高效残差分解网络(efficient residual factorized network, ERFNet)和注意力机制,有效解决了车道线检测中正负样本不平衡的问题,实现了恶劣环境下道路标志和车道线的精准检测。针对混凝土桥梁病害图像语义分割时存在的精度不足、移动端设备计算能力不足等问题,黄彩萍等[121]通过构建多种语义分割模型对混凝土桥梁病害图像进行分割,并引入不同注意力机制进行多角度对比,找到了不同需求下的最佳模型。郭强等[122]基于高分辨率网络模型(high-resolution net version 2, HRNetV2),结合迁移学习及坐标注意力机制提出了一种高寒区隧道衬砌冻害检测方法,平均交并比可达89.25%。
Ministry of Transport of the People’s Republic of China. 2024 statistical communiqué on the development of the transport industry[EB/OL]. (2025-06-12)[2026-02-10]. https://xxgk.mot.gov.cn/2020/jigou/zhghs/202506/t20250610_4170228.html.
Ministry of Transport of the People’s Republic of China. Notice on issuing the action plan for new infrastructure construction in the transport sector (2021-2025): Jiao Gui Hua Fa〔2021〕 No. 82[EB/OL]. (2021-09-23)[2026-02-10]. https://xxgk.mot.gov.cn/2020/jigou/zhghs/202109/t20210923_3619709.html.
Ministry of Transport of the People’s Republic of China, Ministry of Science and Technology of the People’s Republic of China. Notice on issuing the medium-and long-term development plan for science and technology innovation in the transport sector (2021-2035): Jiao Ke Ji Fa〔2022〕 No.11[EB/OL]. (2022-01-24)[2026-02-10]. https://www.gov.cn/zhengce/zhengceku/2022-04/06/content_5683595.htm.
XueYadong, GaoJian, LiYicheng, et al. Optimization of shield tunnel lining defect detection model based on deep learning[J]. Journal of Hunan University (Natural Sciences Edition), 2020, 47(7): 137-146.
LiuDejun, ZhongFei, HuangHongwei, et al. Present status and development trend of diagnosis and treatment of tunnel lining diseases[J]. China Journal of Highway and Transport, 2021, 34(11): 178-199.
RenSong, ZhuQianwen, TuXinyue, et al. Lining disease identification of highway tunnel based on deep learning[J]. Journal of Zhejiang University (Engineering Science Edition), 2022, 56(1): 92-99.
[15]
WangB, HeN, XuF, et al. Visual detection method of tunnel water leakage diseases based on feature enhancement learning[J]. Tunnelling and Underground Space Technology, 2024, 153:106009.
[16]
XuH, WangM, LiuC, et al. Automatic detection of tunnel lining crack based on mobile image acquisition system and deep learning ensemble model[J]. Tunnelling and Underground Space Technology, 2024, 154: 106124.
TanYuanxin, GuoRunhua, ShiPengcheng. Dynamic inspection technology of tunnel deformation based on onboard LIDAR[J]. Urban Mass Transit, 2024, 27(6): 276-280, 285.
ChuXiumin, YanXinping, HaoJing. Design of multi-functional automatic data acquisition vehicle for high-grade highway conditions[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering Edition), 2007(3): 438-441.
ZhuJie, LiLi, LeiQianrong, et al. A review of 3D pavement inspection technology and equipment[J]. Journal of Highway and Transportation Research and Development(Application Technology Edition), 2019, 15(11): 30-33.
[29]
HuangH, SunY, XueY, et al. Inspection equipment study forsubway tunnel defects by grey-scale image processing[J]. Advanced Engineering Informatics, 2017, 32: 188-201.
FuChunshuang. Research on the surface crack and leakage detection system of subway tunnel based on machine vision[D]. Chengdu: Southwest Jiaotong University, 2023.
XiangYang. A lightweight algorithm based on migration fusion learning for image recognition of road tunnel lining cracks[D]. Chongqing: Chongqing Jiaotong University, 2024.
[36]
YuS N, JangJ H, HanC S. Auto inspection system using amobile robot for detecting concrete cracks in a tunnel[J]. Automation in Construction, 2007, 16(3): 255-261.
WangJianfeng, MaJian, MaRonggui, et al. Research on 3D pavement inspection technology[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering Edition), 2010, 34(6): 1202-1205.
[39]
StentS, GherardiR, StengerB, et al. Visual change detection on tunnel linings[J]. Machine Vision and Applications, 2014, 27(3): 319-330.
[40]
HuangH, LiQ, ZhangD. Deep learning based image recognition for crack and leakage defects of metro shield tunnel[J]. Tunnelling and Underground Space Technology, 2018, 77: 166-176.
[41]
LiD, XieQ, GongX, et al. Automatic defect detection of metro tunnel surfaces using a vision-based inspection system[J]. Advanced Engineering Informatics, 2021, 47: 101206.
ShiBo, WangHuichao, YangXiaolei, et al. Design and implementation of rail crossing tunnel disease detection system combined with linear array camera/scanner[J]. Journal of Electronic Mea-surement and Instrumentation, 2024, 38(6): 181-194.
LiuYongfeng, LiChi, HuangBingjie, et al. Application of mobile 3D laser scanning rail inspection vehicle in ellipticity detection of metro shield tunnels[J]. Urban Geotechnical Investigation & Surveying, 2020(6): 119-122.
LiYongbing, GaoChengming, MaYingying, et al. Application of 3D laser scanning technology in deformation monitoring and detection of tunnel[J]. Science Technology and Engineering, 2021, 21(12): 5111-5117.
LuChen, WangKui, ShiYichao, et al. Tunnel cross-section deformation detection based on 3D laser scanning technology[J]. Science Technology and Engineering, 2025, 25(24): 10459-10467.
[54]
JeongS, KimM G, KimS T, et al. Crack inspection in tunnel structures by fusing information from a 3D light detection and ranging and pantilt-zoom camera system[J]. Structures, 2023, 58: 105420.
[55]
MenendezE, VictoresJ G, MonteroR, et al. Tunnel structural inspection and assessment using an autonomous robotic system[J]. Automation in Construction, 2018, 87: 117-126.
[56]
AiQ, YuanY. Rapid acquisition and identification of structural defects of metro tunnel[J]. Sensors, 2019, 19(19): 4278.
[57]
NakamuraS, YamashitaA, InoueF, et al. Inspection test of a tunnel with an inspection vehicle for tunnel lining concrete[J]. Journal of Robotics and Mechatronics, 2019, 31(6): 762-771.
[58]
YasudaT, YamamotoH, EnomotoM, et al. Smart tunnel inspection and assessment using mobile inspection vehicle, non-contact radar and AI[C]// 37th International Symposium on Automation and Robotics in Construction. Kitakyushu: IAARC Publications, 2020: 1373-1379.
[59]
FujinoY, SiringoringoD M. Recent research and development programs for infrastructuresmaintenance, renovation and management in Japan[J]. Structure and Infrastructure Engineering, 2020, 16(1): 3-25.
HuangXi, YaoLianbi, YangLujing. Integrated detection of rail transit building limit based on mobile laser scanning[J]. Bulletin of Surveying and Mapping, 2023(3): 84-88.
[62]
LiaoJ, YueY, ZhangD, et al. Automatic tunnel crack inspection using an efficient mobile imaging module and a lightweight CNN[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(9): 15190-15203.
[63]
SunH, LiuS, ZhongR, et al. Cross-section deformation analysis and visualization of shield tunnel based on mobile tunnel monitoring system[J]. Sensors, 2020, 20(4): 1006.
[64]
SjölanderA, BelloniV, RavanelliR, et al. TACK-an autonomous inspection system for tunnels[C]// Proceedings of the 47th ITA-AITES World Tunnel Congress. Copenhagen: ITA-AITES, 2022: 1-9.
[65]
ShaoW, VijayaranganS, LiC, et al. Stereo visual inertial LiDAR simultaneous localization and mapping[C]// IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Macau: IEEE, 2019: 370-377.
[66]
FilipI, PyoJ, LeeM, et al. Lidar SLAM comparison in a featureless tunnel environment[C]// IEEE Conference on Control, Automation and Systems (ICCAS). Busan: IEEE, 2022: 1648-1653.
QuanMeixiang, PiaoSonghao, LiGuo. An overview of visual SLAM[J]. CAAI Transactions on Intelligent Systems, 2016, 11(6): 768-776.
[69]
DuanX, HuQ, AiM, et al. STATIC-LIO: a sliding window and terrain-assisted dynamic points removal LiDAR inertial odometry[J]. Information Fusion, 2025, 121: 103132.
[70]
SolodarD, KleinI. VIO-DualProNet: visual-inertial odometry with learning based process noise covariance[J]. Engineering Applications of Artificial Intelligence, 2024, 133: 108466.
[71]
HuangG. Visual-inertial navigation: a concise review[C]// IEEE Conference on Robotics and Automation (ICRA). Mont-real: IEEE, 2019: 9572-9582.
[72]
LinH Y, YehM C. Drift-free visual SLAM for mobile robot localization by integrating UWB technology[J]. IEEE Access, 2022, 10: 93636-93645.
WangShilei, GaoYan, QiFalin, et al. Review on inspection technology of railway operation tunnels[J]. Journal of Traffic and Transportation Engineering, 2020, 20(5): 41-57.
[75]
LandE H, McCannJ J. Lightness and retinex theory[J]. Journal of the Optical Society of America, 1971, 61(1): 1-11.
MaLong, MaTengyu, LiuRisheng. The review of low-light image enhancement[J]. Journal of Image and Graphics, 2022, 27(5): 1392-1409.
[80]
WangJ, TanW, MuX, et al. RDGAN: retinex decomposition based adversarial learning for low-light enhancement[C]// IEEE International Conference on Multimedia and Expo (ICME). Shanghai: IEEE, 2019: 1186-1191.
[81]
SunJ, CaoW F, XuZ B, et al. Learning a convolutional neural network for non-uniform motion blur removal[C]// IEEE Confe-rence on Computer Vision and Pattern Recognition (CVPR). Boston: IEEE, 2015: 769-777.
[82]
NahS, KimT H, LeeK M. Deep multi-scale convolutional neural network for dynamic scene deblurring[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu: IEEE, 2017: 257-265.
[83]
KupynO, BudzanV, MykhailychM, et al. DeblurGAN: blind motion deblurring using conditional adversarial networks[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 8183-8192.
HuangZhengyuan, XieWeicheng, HuangHuaru, et al. Overview of motion deblurring techniques for images[J]. Computer Engineering and Applications, 2020, 56(24): 28-34.
[86]
ZamirS W, AroraA, KhanS, et al. Restormer: efficient transformer for high-resolution image restoration[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans: IEEE, 2022: 5718-5729.
[87]
AbdulkhaleqA H, El AbbadiN K, Al HassaniS A. A survey on blind de-blurring of digital image[J]. Iraqi Journal of Science, 2022: 338-352.
[88]
ChoS J, JiS W, HongJ P, et al. Rethinking coarse-to-fine approach in single image deblurring[C]// IEEE/CVF International Conference on Computer Vision (ICCV). Montreal: IEEE, 2021: 4621-4630.
[89]
MaY, KongF, LuD, et al. An intelligent method for apparent water leakage segmentation of tunnel lining images considering motion blurs[J]. Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards, 2025, 19(4): 907-923.
BaoXiaomin, WangSiqi. Survey of object detection algorithm based on deep learning[J]. Transducer and Microsystem Techno-logies, 2022, 41(4): 5-9.
[94]
GirshickR, DonahueJ, DarrellT, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Columbus: IEEE, 2014: 580-587.
[95]
RenS, HeK, GirshickR, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[96]
DengJ, LuY, LeeV C S. Concrete crack detection with handwriting script interferences using faster region-based convolutional neural network[J]. Computer-Aided Civil and Infrastructure Engineering, 2020, 35(4): 373-388.
[97]
HanP, YangX. Deep learning-enabled automatic detection of bridges for promoting transportation surveillance under different imaging conditions[J]. Journal of Advanced Transportation, 2022, 2022(1): 6932040.
CuiGuangyan, WangYanhui, XuJie, et al. Automatic detection of discrete entity objects in tunnel lining based on improved Faster R-CNN[J]. Journal of the China Railway Society, 2024, 46(2): 171-180.
[102]
RedmonJ, DivvalaS, GirshickR, et al. You only look once: unified, real-time object detection[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas: IEEE, 2016: 779-788.
[103]
RedmonJ, FarhadiA. YOLO9000: better, faster, stronger[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu: IEEE, 2017: 6517-6525.
BochkovskiyA, WangC Y, LiaoH Y M. YOLOv4:optimal speed and accuracy of object detection[EB/OL]. (2020-04-23)[2026-02-10]. https://arxiv.org/abs/2004.10934.
[106]
LiC, LiL, JiangH, et al. YOLOv6:a single-stage object detection framework for industrial applications[EB/OL]. (2022-09-07)[2026-02-10]. https://arxiv.org/abs/2209.02976.
[107]
WangC Y, BochkovskiyA, LiaoH Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver: IEEE, 2023: 7464-7475.
[108]
HussainM. YOLO-v1 to YOLO-v8, the rise of YOLO and its complementary nature toward digital manufacturing and industrial defect detection[J]. Machines, 2023, 11(7): 677.
[109]
WangC Y, YehI H, LiaoH Y M. YOLOv9:learning what you want to learn using programmable gradient information[EB/OL]. (2024-02-29)[2026-02-10]. https://arxiv.org/abs/2402.13616.
LeiM, LiS, WuY, et al. YOLOv13:real-time object detection with hypergraph-enhanced adaptive visual perception[EB/OL]. (2025-06-21)[2026-02-10]. https://arxiv.org/abs/2506.17733.
[115]
WanF, SunC, HeH, et al. YOLO-LRDD: a lightweight method for road damage detection based on improved YOLOv5s[J]. EURASIP Journal on Advances in Signal Processing, 2022(1): 98.
[116]
SongW, SuandiS A. TSR-YOLO: a Chinese traffic sign recognition algorithm for intelligent vehicles in complex scenes[J]. Sensors, 2023, 23(2): 749.
PengHongxing, YuanChang, KeWeiye, et al. Vehicle and personnel detection in highway tunnels based on improved YOLOv5[J]. Science Technology and Engineering, 2024, 24(6): 2453-2461.
HuangJipeng, ShiYinghuan, GaoYang. Multi-scale Faster-RCNN algorithm for small object detection[J]. Journal of Compu-ter Research and Development, 2019, 56(2): 319-327.
[125]
CaiZ, VasconcelosN. Cascade R-CNN: high quality object detection and instance segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(5): 1483-1498.
FuJinyi, ZhangZijia, SunWei, et al. Improved YOLOv8 small target detection algorithm in aerial images[J]. Computer Engineering and Applications, 2024, 60(6): 100-109.
[132]
ChenJ, XuX, JeonG, et al. WLR-net: an improved YOLO-V7 with edge constraints and attention mechanism for water leakage recognition in the tunnel[J]. IEEE Transactions on Emerging Topics in Computational Intelligence, 2024, 8(4): 3105-3116.
[133]
ZhouZ, ZhangJ, GongC. Automatic detection method of tunnel lining multi-defects via an enhanced you only look once network[J]. Computer-Aided Civil and Infrastructure Engineering, 2022, 37(6): 762-780.
LiuJie, JinJide, ZhengQingxiang. Night vehicle detection method based on improving Mask RCNN[J]. Journal of Transport Information and Safety, 2023, 41(2): 59-66.
WangYunpeng, WuQiong, SongDewang, et al. A survey of autonomous driving datasets and 3D object perception methods[J]. Artificial Intelligence, 2023(5): 31-47.
ZhuangFuzhen, LuoPing, HeQing, et al. Survey on transfer learning research[J]. Journal of Software, 2015, 26(1): 26-39.
[140]
SantoroA, BartunovS, BotvinickM, et al. Meta-learning with memory-augmented neural networks[C]// IEEE Conference on Machine Learning. New York: PMLR, 2016: 1842-1850.
[141]
GoodfellowI, Pouget-AbadieJ, MirzaM, et al. Generative adversarial nets[C]// Advances in Neural Information Processing Systems 27. Montreal: Curran Associates, Inc., 2014: 2672-2680.
[142]
ZhangG, CuiK, HungT Y, et al. Defect-GAN: high-fidelity defect synthesis for automated defect inspection[C]// IEEE Winter Conference on Applications of Computer Vision (WACV). Waikoloa: IEEE, 2021: 2523-2533.
LiuYing, LeiYanbo, FanJiulun, et al. Survey on image classification technology based on small sample learning[J]. Acta Automatica Sinica, 2021, 47(2): 297-315.
ZhouZhong, YanLongbin, ZhangJunjie, et al. Review and prospect of intelligent identification of apparent diseases in highway tunnels based on deep learning[J]. China Civil Engineering Journal, 2022, 55(S2): 38-48.
[147]
ZhaoG, WangT, YeJ. Anisotropic clustering on surfaces for crack extraction[J]. Machine Vision and Applications, 2015, 26(5): 675-688.
[148]
ChaY, ChoiW, SuhG, et al. Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types[J]. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(9): 731-747.
[149]
MaedaH, SekimotoY, SetoT, et al. Road damage detection and classification using deep neural networks with smartphone images[J]. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(12): 1127-1141.
[150]
XueY, LiY. A fast detection method via region-based fully convolutional neural networks for shield tunnel lining defects[J]. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(8): 638-654.
ChengWangliu, RenYangxun, NiXiufeng, et al. Defect recognition of vibration dampers based on improved Cascade R-CNN network model[J]. Journal of Anhui University (Natural Science Edition), 2022, 46(5): 64-70.
LiuXiaolei, WuGuoqun, KanZhe. Research on defect detection method of coal mine wire rope based on deep learning[J]. Coal Engineering, 2023, 55(11): 148-153.
[155]
邱军富. 道路交通标识的缺损检测算法研究[D]. 南京: 东南大学, 2023.
[156]
QiuJunfu. Research on defect detection algorithm of road traffic sign[D]. Nanjing: Southeast University, 2023.
[157]
AcikgozH, KorkmazD. MSRConvNet: classification of railway track defects using multi-scale residual convolutional neural network[J]. Engineering Applications of Artificial Intelligence, 2023, 121: 105965.
GouJunnian, DuSusu, WangShiduo, et al. Insulator self-explosion detection in transmission line based on CenterNet fusing lightweight features[J]. Journal of Beijing University of Aeronautics and Astronautics, 2024, 50(7): 2161-2171.
Shanxi Intelligent Transportation Laboratory Co., Ltd., TongjiUniversity. A tunnel visual facility identification system and its storage medium: 202510440388.5[P]. 2025-07-25.
[162]
YunH, ParkD. Efficient object detection based on masking semantic segmentation region for lightweight embedded processors[J]. Sensors, 2022, 22(22): 8890.
[163]
LongJ, ShelhamerE, DarrellT, et al. Fully convolutional networks for semantic segmentation[C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston: IEEE, 2015: 3431-3440.
[164]
RonnebergerO, FischerP, BroxT. U-net: convolutional networks for biomedical image segmentation[M]// NavabN, HorneggerJ, WellsW M, et al. Medical Image Computingand Computer-assisted Intervention. Cham: Springer International Publishing, 2015.
[165]
ChenL C, PapandreouG, KokkinosI, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[166]
BadrinarayananV, KendallA, CipollaR. SegNet: a deep convo-lutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495.
[167]
HaltakovV, MayrJ, UngerC, et al. Semantic segmentation based traffic light detection at day and at night[M]// GallJ, GehlerP, LeibeB. Pattern Recognition: Vol. 9358. Cham:Springer International Publishing, 2015.
[168]
LeeJ S, HwangS H, ChoiI Y, et al. Estimation of crack width based on shape-sensitive kernels and semantic segmentation[J]. Structural Control and Health Monitoring, 2020, 27: e2504.
[169]
ZhangL, ShenJ, ZhuB. A research on an improved unet-based concrete crack detection algorithm[J]. Structural Health Monitoring, 2021, 20(4): 1864-1879.
[170]
ChenT, ChenA. Road sign recognition method based on segmentation and attention mechanism[J]. Mobile Information Systems, 2022, 2022: 1-11.
GuoQiang, CheBowen, BaoWeixing, et al. Freezing damage detection method of tunnel lining in cold regions based on improved HRNetV2[J]. Science Technology and Engineering, 2024, 24(25): 10956-10963.
LiaoYanna, DouDanyang. Design and research of bridge cracks detection method based on Mask RCNN[J]. Journal of Applied Optics, 2022, 43(1): 100-105, 118.
WangBuyun, LiHongwei, ZhaoShan. Road side multi-object recognition by integrating point cloud and panoramic image[J]. Bulletin of Surveying and Mapping, 2023(10): 40-46.