臺灣在橋梁逐年老舊的情況下，對於興建後的長期監測與例行性維護已為管理單位的年度執行重點。其中又以橋梁底版及下部構件檢測最為困難，實因橋梁位置常距離河床面遠，檢測員難在高空近距離人工目視檢測，易衍生勞工安全危害，致使檢測效率低落。近年來橋梁管理單位在例行巡檢需求量逐年攀升，另需於橋梁的使用年限內編列可觀的維修經費，已造成沉痾的財務負擔。經審閱檢視多家施工廠商的歷史估算表後，發現編列之工程項目以及維修單價尚無制式規範，遑論自動化的推估劣化維修經費。爰此，如何合理預測橋梁生命週期內的例行維修成本，可讓相關管理單位編列合理預算，或提供承包商劣化型式及修復區域，成為投標文件的需求背景說明，為本研究的主要目的。經現地遙控操作無人飛行載具採集臺灣中部以北之13座複合式橋梁、18座鋼構橋梁以及30座鋼筋混凝土橋梁的劣化影像，統計發現橋梁底版及下部元件之常見劣化態樣可分為混凝土破裂、混凝土裂縫、混凝土蜂窩、鋼筋裸露鏽蝕、鋼鈑脫漆、鋼鈑鏽蝕以及螺栓鏽蝕，故本研究以此七種劣化類別作為後續YOLOv7實例分割模型之辨識目標，並使用PWO-Lite元啟發式優化演算法找到最佳的超參數組合，以提升模型的精確度。首先，藉由計算機電腦視覺圖像分割技術框列劣化範圍，以橋梁劣化態樣對應之維修工法，推估劣化元件維修的總經費成本；最後，建構無人機檢測影像劣化維護成本估算系統，研究成果希冀提供予橋梁管理單位或施工廠商作為日後橋梁檢修之決策資訊。

In Taiwan, as bridges continue to age, the focus of management authorities has shifted toward long-term monitoring and routine maintenance of newly constructed bridges. Among these, the inspection of bridge decks and substructure components presents significant challenges. Due to the often remote location of bridges from riverbeds, conducting close-range visual inspections at great heights is difficult for inspectors, resulting in low inspection efficiency and potential risks to worker safety. In recent years, the demand for routine inspections has steadily increased, requiring substantial budget allocations for maintenance throughout the lifespan of bridges, imposing a financial burden. After reviewing estimation tables from various construction firms, it was discovered that there are no standardized specifications for project listings and maintenance unit prices, let alone automated estimation of repair costs for deterioration. Hence, the main objective of this study is to develop a reliable method for predicting routine maintenance costs throughout the lifecycle of bridges. This will enable relevant management authorities to allocate reasonable budgets, provide background explanations for bidding documents concerning types of deterioration and repair areas to contractors, and serve as a foundation for decision-making in bridge inspections. By utilizing remotely operated unmanned aerial vehicles (UAVs), a collection of deterioration images was gathered from 13 composite bridges, 18 steel bridges, and 30 reinforced concrete bridges in the central and northern regions of Taiwan. Statistical analysis revealed common deterioration patterns in bridge decks and substructure elements, including concrete cracking, concrete spalling, concrete honeycombing, exposed and corroded reinforcement bars, steel plate peeling, rusting steel plates, and corroded bolts. Consequently, these seven deterioration categories were selected as the recognition targets for subsequent YOLOv7 instance segmentation models. The PWO-Lite metaheuristic optimization algorithm was employed to find the optimal set of hyperparameters, thereby enhancing the accuracy of the model. The proposed approach involves utilizing computer vision image segmentation techniques to identify deterioration areas and estimate the total cost of repairing deteriorated components based on corresponding maintenance methods. Finally, an unmanned aerial vehicle inspection image deterioration maintenance cost estimation system will be established. The research outcomes aim to provide decision-making information for bridge management authorities and construction firms in future bridge inspections.

[1] N. J. Yau and Y. H. Chuang, "Analyzing Taiwan Bridge Management System for decision making in bridge maintenance: A big data approach," in 2015 10th International Joint Conference on Software Technologies (ICSOFT), 20-22 July 2015 2015, vol. 1, pp. 1-6.
[2] 宋裕祺, 張國鎮, 陳俊仲, 李政寬, 洪曉慧, and 蘇進國, "橋梁監測與結構健康診斷," (in 繁體中文), 土木水利, vol. 45, no. 5, pp. 22-29, 2018, doi: 10.6653/MoCICHE.201810_45(5).0003.
[3] C.-Y. Liu and J.-S. Chou, "Bayesian-optimized deep learning model to segment deterioration patterns underneath bridge decks photographed by unmanned aerial vehicle," Automation in Construction, vol. 146, p. 104666, 2023/02/01/ 2023, doi: https://doi.org/10.1016/j.autcon.2022.104666.
[4] H. Wan, L. Gao, Z. Yuan, H. Qu, Q. Sun, H. Cheng, and R. Wang, "A novel transformer model for surface damage detection and cognition of concrete bridges," Expert Systems with Applications, vol. 213, p. 119019, 2023/03/01/ 2023, doi: https://doi.org/10.1016/j.eswa.2022.119019.
[5] L. Chen, W. Chen, L. Wang, C. Zhai, X. Hu, L. Sun, Y. Tian, X. Huang, and L. Jiang, "Convolutional neural networks (CNNs)-based multi-category damage detection and recognition of high-speed rail (HSR) reinforced concrete (RC) bridges using test images," Engineering Structures, vol. 276, p. 115306, 2023/02/01/ 2023, doi: https://doi.org/10.1016/j.engstruct.2022.115306.
[6] L. Hebbache, D. Amirkhani, M. S. Allili, N. Hammouche, and J.-F. Lapointe, "Leveraging Saliency in Single-Stage Multi-Label Concrete Defect Detection Using Unmanned Aerial Vehicle Imagery," Remote Sensing, vol. 15, no. 5, doi: 10.3390/rs15051218.
[7] H. Zhang, W. Qin, Z. Zhou, P. Zhu, and W. Du, "Piezomagnetoelastic energy harvesting from bridge vibrations using bi-stable characteristics," Energy, vol. 263, p. 125859, 2023/01/15/ 2023, doi: https://doi.org/10.1016/j.energy.2022.125859.
[8] J. Yu, W. Zhou, L. Jiang, K. Peng, and L. Zu, "Train effect on the vibration behavior of high-speed railway track-bridge system subjected to seismic excitation," Soil Dynamics and Earthquake Engineering, vol. 172, p. 108049, 2023/09/01/ 2023, doi: https://doi.org/10.1016/j.soildyn.2023.108049.
[9] S. S. Khedmatgozar Dolati, P. Malla, J. D. Ortiz, A. Mehrabi, and A. Nanni, "Identifying NDT methods for damage detection in concrete elements reinforced or strengthened with FRP," Engineering Structures, vol. 287, p. 116155, 2023/07/15/ 2023, doi: https://doi.org/10.1016/j.engstruct.2023.116155.
[10] S. Kashif Ur Rehman, Z. Ibrahim, S. A. Memon, and M. Jameel, "Nondestructive test methods for concrete bridges: A review," Construction and Building Materials, vol. 107, pp. 58-86, 2016/03/15/ 2016, doi: https://doi.org/10.1016/j.conbuildmat.2015.12.011.
[11] B. F. Spencer, V. Hoskere, and Y. Narazaki, "Advances in Computer Vision-Based Civil Infrastructure Inspection and Monitoring," Engineering, vol. 5, no. 2, pp. 199-222, 2019/04/01/ 2019, doi: https://doi.org/10.1016/j.eng.2018.11.030.
[12] M. Abbas and M. Shafiee, "Structural health monitoring (SHM) and determination of surface defects in large metallic structures using ultrasonic guided waves," Sensors, vol. 18, no. 11, p. 3958, 2018.
[13] R. Jones, "The ultrasonic testing of concrete," Ultrasonics, vol. 1, no. 2, pp. 78-82, 1963.
[14] X. Yan, D. Zhang, S. Pan, E. Zhang, and W. Gao, "Online nondestructive testing for fine steel wire rope in electromagnetic interference environment," NDT & E International, vol. 92, pp. 75-81, 2017.
[15] J.-W. Kim and S. Park, "Magnetic flux leakage sensing and artificial neural network pattern recognition-based automated damage detection and quantification for wire rope non-destructive evaluation," Sensors, vol. 18, no. 1, p. 109, 2018.
[16] A. Vakhguelt, S. D. Kapayeva, and M. J. Bergander, "Combination Non-Destructive Test (NDT) Method for Early Damage Detection and Condition Assessment of Boiler Tubes," Procedia Engineering, vol. 188, pp. 125-132, 2017/01/01/ 2017, doi: https://doi.org/10.1016/j.proeng.2017.04.465.
[17] R. Gupta, D. Mitchell, J. Blanche, S. Harper, W. Tang, K. Pancholi, L. Baines, D. G. Bucknall, and D. Flynn, "A review of sensing technologies for non-destructive evaluation of structural composite materials," Journal of Composites Science, vol. 5, no. 12, p. 319, 2021.
[18] G. Li, S. He, Y. Ju, and K. Du, "Long-distance precision inspection method for bridge cracks with image processing," Automation in Construction, vol. 41, pp. 83-95, 2014/05/01/ 2014, doi: https://doi.org/10.1016/j.autcon.2013.10.021.
[19] E. Peli, "Feature detection algorithm based on a visual system model," Proceedings of the IEEE, vol. 90, no. 1, pp. 78-93, 2002.
[20] M. Banerjee and M. K. Kundu, "Edge based features for content based image retrieval," Pattern Recognition, vol. 36, no. 11, pp. 2649-2661, 2003.
[21] Y. Huachao, Z. Shubi, and W. Yongbo, "Robust and precise registration of oblique images based on scale-invariant feature transformation algorithm," IEEE Geoscience and Remote Sensing Letters, vol. 9, no. 4, pp. 783-787, 2012.
[22] L. Vincent, "Morphological algorithms," in Mathematical Morphology in Image Processing: CRC Press, 2018, pp. 255-288.
[23] F. Provost, C. Hibert, and J. P. Malet, "Automatic classification of endogenous landslide seismicity using the Random Forest supervised classifier," Geophysical Research Letters, vol. 44, no. 1, pp. 113-120, 2017.
[24] G. M. Hadjidemetriou, P. A. Vela, and S. E. Christodoulou, "Automated pavement patch detection and quantification using support vector machines," Journal of Computing in Civil Engineering, vol. 32, no. 1, p. 04017073, 2018.
[25] K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," arXiv preprint arXiv:1409.1556, 2014.
[26] C.-C. Hsu and C.-W. Lin, "Unsupervised convolutional neural networks for large-scale image clustering," in 2017 IEEE International Conference on Image Processing (ICIP), 2017: IEEE, pp. 390-394.
[27] X. Chai, Z. Liu, F. Yin, Z. Liu, and X. Chen, "Two streams recurrent neural networks for large-scale continuous gesture recognition," in 2016 23rd international conference on pattern recognition (ICPR), 2016: IEEE, pp. 31-36.
[28] P. Domingos, "A few useful things to know about machine learning," Communications of the ACM, vol. 55, no. 10, pp. 78-87, 2012.
[29] Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," nature, vol. 521, no. 7553, pp. 436-444, 2015.
[30] X. Peng, X. Zhong, C. Zhao, A. Chen, and T. Zhang, "A UAV-based machine vision method for bridge crack recognition and width quantification through hybrid feature learning," Construction and Building Materials, vol. 299, p. 123896, 2021/09/13/ 2021, doi: https://doi.org/10.1016/j.conbuildmat.2021.123896.
[31] H. Su, S. Wei, M. Yan, C. Wang, J. Shi, and X. Zhang, "Object detection and instance segmentation in remote sensing imagery based on precise mask R-CNN," in IGARSS 2019-2019 IEEE International Geoscience and Remote Sensing Symposium, 2019: IEEE, pp. 1454-1457.
[32] N. Dong, Y. Zhang, M. Ding, and G. H. Lee, "Bridging non co-occurrence with unlabeled in-the-wild data for incremental object detection," Advances in Neural Information Processing Systems, vol. 34, pp. 30492-30503, 2021.
[33] H. Fu, D. Meng, W. Li, and Y. Wang, "Bridge crack semantic segmentation based on improved Deeplabv3+," Journal of Marine Science and Engineering, vol. 9, no. 6, p. 671, 2021.
[34] B. Kim and S. Cho, "Automated multiple concrete damage detection using instance segmentation deep learning model," Applied Sciences, vol. 10, no. 22, p. 8008, 2020.
[35] X. Qian, C. Li, W. Wang, X. Yao, and G. Cheng, "Semantic segmentation guided pseudo label mining and instance re-detection for weakly supervised object detection in remote sensing images," International Journal of Applied Earth Observation and Geoinformation, vol. 119, p. 103301, 2023/05/01/ 2023, doi: https://doi.org/10.1016/j.jag.2023.103301.
[36] H. R. Zhang, D. S. Xu, D. Y. Cheng, X. L. Meng, G. Xu, W. Liu, and T. Wang, "An Improved Lightweight Yolo-Fastest V2 for Engineering Vehicle Recognition Fusing Location Enhancement and Adaptive Label Assignment," IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, vol. 16, pp. 2450-2461, 2023, doi: 10.1109/JSTARS.2023.3249216.
[37] C.-Y. Wang, A. Bochkovskiy, and H.-Y. M. Liao, "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors," arXiv preprint arXiv:2207.02696, 2022.
[38] R. Santos, D. Ribeiro, P. Lopes, R. Cabral, and R. Calçada, "Detection of exposed steel rebars based on deep-learning techniques and unmanned aerial vehicles," Automation in Construction, vol. 139, p. 104324, 2022/07/01/ 2022, doi: https://doi.org/10.1016/j.autcon.2022.104324.
[39] V. Pham, D. Nguyen, and C. Donan, "Road Damage Detection and Classification with YOLOv7," in 2022 IEEE International Conference on Big Data (Big Data), 17-20 Dec. 2022 2022, pp. 6416-6423, doi: 10.1109/BigData55660.2022.10020856.
[40] H. Bian, Y. Liu, L. Shi, Z. Lin, M. Huang, J. Zhang, G. Weng, C. Zhang, and M. Gao, "Detection Method of Helmet Wearing Based on UAV Images and Yolov7," in 2023 IEEE 6th Information Technology,Networking,Electronic and Automation Control Conference (ITNEC), 24-26 Feb. 2023 2023, vol. 6, pp. 1633-1640, doi: 10.1109/ITNEC56291.2023.10082536.
[41] X. Duan, H. Chen, H. Lou, L. Bi, Y. Zhang, and H. Liu, "A more accurate mask detection algorithm based on Nao robot platform and YOLOv7," in 2023 IEEE 3rd International Conference on Power, Electronics and Computer Applications (ICPECA), 29-31 Jan. 2023 2023, pp. 1295-1299, doi: 10.1109/ICPECA56706.2023.10076188.
[42] S. Hudson, R. Carmichael III, L. Moser, W. Hudson, and W. Wilkes, "Bridge management systems," NCHRP Report, no. 300, 1987.
[43] V. Gattulli and L. Chiaramonte, "Condition assessment by visual inspection for a bridge management system," Computer‐Aided Civil and Infrastructure Engineering, vol. 20, no. 2, pp. 95-107, 2005.
[44] A. Miyamoto, K. Kawamura, and H. Nakamura, "Bridge management system and maintenance optimization for existing bridges," Computer‐Aided Civil and Infrastructure Engineering, vol. 15, no. 1, pp. 45-55, 2000.
[45] F. Akgül and M. Frangopol Dan, "Computational Platform for Predicting Lifetime System Reliability Profiles for Different Structure Types in a Network," Journal of Computing in Civil Engineering, vol. 18, no. 2, pp. 92-104, 2004/04/01 2004, doi: 10.1061/(ASCE)0887-3801(2004)18:2(92).
[46] B. Kövesdi, D. Kollár, L. Dunai, and A. Horváth, "Reliability analysis-based investigation of the historical Széchenyi Chain Bridge deck system," Results in Engineering, vol. 15, p. 100555, 2022/09/01/ 2022, doi: https://doi.org/10.1016/j.rineng.2022.100555.
[47] C.-C. Comisu, N. Taranu, G. Boaca, and M.-C. Scutaru, "Structural health monitoring system of bridges," Procedia Engineering, vol. 199, pp. 2054-2059, 2017/01/01/ 2017, doi: https://doi.org/10.1016/j.proeng.2017.09.472.
[48] A. Salzano, A. Miano, M. Nicolella, and A. Prota, "The structural maintenance process of existing bridges integrated in BIM tools," Materials Today: Proceedings, vol. 85, pp. 9-13, 2023/01/01/ 2023, doi: https://doi.org/10.1016/j.matpr.2023.05.244.
[49] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 27-30 June 2016 2016, pp. 779-788, doi: 10.1109/CVPR.2016.91.
[50] J. Redmon and A. Farhadi, "YOLO9000: Better, Faster, Stronger," in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 21-26 July 2017 2017, pp. 6517-6525, doi: 10.1109/CVPR.2017.690.
[51] J. Redmon and A. Farhadi, "Yolov3: An incremental improvement," arXiv preprint arXiv:1804.02767, 2018.
[52] A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao, "Yolov4: Optimal speed and accuracy of object detection," arXiv preprint arXiv:2004.10934, 2020.
[53] G. R. Jocher, A. Stoken, J. Borovec, NanoCode, A. Chaurasia, TaoXie, L. Changyu, Abhiram, Laughing, tkianai, yxNONG, A. Hogan, lorenzomammana, AlexWang, J. Hájek, L. Diaconu, Marc, Y. Kwon, Oleg, wanghaoyang, Y. Defretin, A. Lohia, m. ah, B. Milanko, B. Fineran, D. P. Khromov, D. Yiwei, Doug, Durgesh, and F. Ingham, "ultralytics/yolov5: v5.0 - YOLOv5-P6 1280 models, AWS, Supervise.ly and YouTube integrations," 2021.
[54] C. Li, L. Li, H. Jiang, K. Weng, Y. Geng, L. Li, Z. Ke, Q. Li, M. Cheng, and W. Nie, "YOLOv6: A single-stage object detection framework for industrial applications," arXiv preprint arXiv:2209.02976, 2022.
[55] M. Yasir, L. Zhan, S. Liu, J. Wan, M. S. Hossain, A. T. Isiacik Colak, M. Liu, Q. U. Islam, S. Raza Mehdi, and Q. Yang, "Instance segmentation ship detection based on improved Yolov7 using complex background SAR images," (in English), Frontiers in Marine Science, Original Research vol. 10, 2023-May-01 2023, doi: 10.3389/fmars.2023.1113669.
[56] X. Ding, "RepVGG: Making VGGstyle convnets great again," 2021, doi: https://github.com/DingXiaoH/RepVGG.
[57] J.-S. Chou and C.-Y. Liu, "Pilgrimage walk optimization: Bobee culture-inspired algorithm for identification of bridge deterioration," Automation in Construction, 2023. R1 review.
[58] L. Xu, W. Yan, and J. Ji, "The research of a novel WOG-YOLO algorithm for autonomous driving object detection," Scientific Reports, vol. 13, no. 1, p. 3699, 2023/03/06 2023, doi: https://doi.org/10.1038/s41598-023-30409-1.
[59] J.-S. Chou and C.-Y. Liu, "Tiny meta-heuristic optimized deep learning edge computing system platform for deterioration detection of concrete underneath bridge decks. ," 2023.
[60] A. Kumar, "Hold-out Method for Training Machine Learning Models," 2020/12/22 2020, doi: https://vitalflux.com/hold-out-method-for-training-machine-learning-model/.
[61] K. M. Ting, "Confusion Matrix," in Encyclopedia of Machine Learning, C. Sammut and G. I. Webb Eds. Boston, MA: Springer US, 2010, pp. 209-209.
[62] H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, and S. Savarese, "Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression," in 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA, 15-20 June 2019 2019, pp. 658-666, doi: 10.1109/CVPR.2019.00075. [Online]. Available: https://ieeexplore.ieee.org/document/8953982
[63] F. Pagani, M. Dell'Amico, and D. Balzarotti, "Beyond precision and recall: understanding uses (and misuses) of similarity hashes in binary analysis," in Proceedings of the Eighth ACM Conference on Data and Application Security and Privacy, 2018, pp. 354-365.
[64] J. Keilwagen, I. Grosse, and J. Grau, "Area under precision-recall curves for weighted and unweighted data," (in eng), PLoS One, vol. 9, no. 3, p. e92209, 2014, doi: 10.1371/journal.pone.0092209.
[65] A. Torralba, B. C. Russell, and J. Yuen, "LabelMe: Online Image Annotation and Applications," Proceedings of the IEEE, vol. 98, no. 8, pp. 1467-1484, 2010, doi: 10.1109/JPROC.2010.2050290.
[66] Aleju and A. Jung, "imgaug, Image augmentation for machine learning experiments.," 2017, doi: https://github.com/aleju/imgaug.
[67] Wang, C.-Y. a. Liao, H.-Y. Mark, and I.-H. Yeh, "Designing Network Design Strategies Through Gradient Path Analysis," arXiv preprint arXiv:2211.04800, 2022, doi: https://github.com/WongKinYiu/yolov7/tree/u7.
[68] I. Banerjee, B. Nguyen, V. Garousi, and A. Memon, "Graphical user interface (GUI) testing: Systematic mapping and repository," Information and Software Technology, vol. 55, no. 10, pp. 1679-1694, 2013/10/01/ 2013, doi: https://doi.org/10.1016/j.infsof.2013.03.004.