提升河川疏濬工區智慧化，建立自動化監控系統，降低人為影響因素，可有效嚇阻及減低犯罪機率，並同時減輕人員工作負擔。智慧工地為運用人工智慧、傳感技術等科技以提升工作效率、降低人力需求、減少錯誤發生，乃全球工業4.0致力追求目標。由於台灣河川疏濬工程領域尚未全面拓展，故本研究規劃設計一疏濬工程智慧工地，依照不同區域之工作性質設計自動化流程。同時，從工區管理、設備維護、災害預防等層面進行考量，輔助智慧工地之規劃，使相關工作得以自動化執行。本研究初期以管制站工作自動化為目標，首先針對河川疏濬場域運輸卡車進入管制站時，以工區監控設備拍攝砂石車影像，以電腦視覺取得影像後，採用深度學習技術YOLOv3執行車牌位置檢測，其平均精度均值(Mean Aveage Precision, mAP)達97.14，速度0.03秒/張。截取框選的車牌圖像另存並輸入至車牌字符數量分類模型，以CNN-L3、SRCS、VGG16神經網路結構判別車牌字符數量類別(6碼、7碼或8碼)，最佳模型CNN-L3的分類正確率達99.90%，速度為0.0315秒/張。再依分類結果將車牌影像傳至對應字符數量之識別字元模型，最佳模型CNN-L3正確率高於97.80%，速度0.0624-0.0781秒/張。整體辨識率93.73%，單碼識別率97.59%，速度0.3271秒/張。初期研發成果有利於疏濬工區的智慧化發展，如(1)藉由自動化監控系統辨識砂石車車牌，雲端資訊管控進出工地車輛(2)協助工區管制站人員每日辨識上百輛砂石車及受工區塵土、砂石附著的車牌，提升識別速度，減少人為錯誤產生；(3)嚇阻不肖業者不法行為，避免相關同仁受暴力恐嚇威脅，降低犯罪風險。上述貢獻可減少人力、時間等耗費，且有效降低不法份子犯罪機會，並使第一線人員工作更為透明化、自動化、效率化。車牌辨識成果為管制站作業自動化的第一步，藉由後續系統功能持續開發，可使疏濬工程規劃與控制朝智慧工地目標邁進。

To this end, raising the intelligence of the dredging area, establishing an automated monitoring system, and reducing human factors can effectively deter and reduce the probability of crime, and at the same time reduce the workload of personnel. Smart site is a global industry 4.0 dedicated to the pursuit of goals by using artificial intelligence, sensor technology and other technologies to improve work efficiency, reduce manpower requirements, and reduce errors. This concept has not yet been fully expanded in the field of river dredging engineering in Taiwan. This study designs a smart construction site for dredging engineering. The automation process is designed according to the goal of the work in different areas. At the same time, it is considered from the construction site management, equipment maintenance, disaster prevention to assist the planning of smart construction sites, so that related work can be automatically executed. The initial purpose was to automate the work of the control station. First of all, when the truck enters the control station in the river dredging field, the image of the truck is taken with the monitoring equipment of the site, and the image is obtained by computer vision. The license plate position in the image is detected with the YOLOv3 model. Its mean aveage precision is 97.14and speed is 0.03 seconds/sheet. The cropped license plate image is input into the license plate character number classification model. Use CNN-L3, SRCS, VGG16 neural network structure to identify the number of license plate characters on the license plate (6 characters, 7 characters or 8 characters). The best model is C-CNN-L3 that accuracy is 99.90% and speed is .0.0315 seconds/sheet. According to the C-CNN-L3 classification result, the license plate image is then transmitted to the corresponding character number recognition character model to predict the license plate number. The best model is R-CNN-L3 that the single character recognition rates are all higher than 97.80%, and the speed is between 0.0624 to 0.0781 seconds/sheet. Finally, the best model at each stage is integrated into a real-time license plate recognition system. The overall recognition rate is 93.73%; the single character recognition rate is 97.59%, and the speed was 0.3271 seconds/sheet. This achievement is conducive to the intelligent development of the dredging area, in the following ways:
(1) identifying truck license plates of through the automatic monitoring system, and controlling the trucks entering and leaving the site in a fair and equitable manner;
(2) assisting the personnel at the control station to identify the license plate with excess dust, sand and stone, thus improving the recognition speed and reducing the occurrence of human error;
(3) deterring unscrupulous industry operators from illegal behavior, thus avoiding the threat of violence and intimidation of related colleagues, and reducing the risk of crime.
These factors can reduce labor and time, and effectively reduce the opportunities for criminals to commit crimes, and make the work more transparent, automated and efficient. The achievement of license plate recognition is the first step in automating the control station operation. The continuous development of subsequent system functions can move the dredging project towards the goal of a smart construction site.

[1] 陳志鵬, "盜採河川砂石犯罪行為之研究—以南投縣為例," 碩士, 犯罪防治所, 國立中正大學, 2005.
[2] 林洸鍇, "河川砂石濫盜採問題與對策之研究," 碩士, 犯罪防治所, 國立中正大學, 2008.
[3] 陳建伊, "河川暨水資源警察設置之可行性研究," 碩士, 犯罪防治所, 國立中正大學, 2007.
[4] 邱柏嘉, "治安策略之決策、執行與評估─以雲林縣警察局取締盜採砂石為例," 碩士, 行政警察研究所, 中央警察大學, 2003.
[5] G. Hearn, Bridge inspection practices. Transportation Research Board, 2007.
[6] S. Agnisarman, S. Lopes, K. Chalil Madathil, K. Piratla, and A. Gramopadhye, "A survey of automation-enabled human-in-the-loop systems for infrastructure visual inspection," Automation in Construction, vol. 97, pp. 52-76, 2019/01/01/ 2019.
[7] Q. Chen, B. García de Soto, and B. T. Adey, "Construction automation: Research areas, industry concerns and suggestions for advancement," Automation in Construction, vol. 94, pp. 22-38, 2018/10/01/ 2018.
[8] R. G. Lins, S. N. Givigi, A. D. M. Freitas, and A. Beaulieu, "Autonomous Robot System for Inspection of Defects in Civil Infrastructures," IEEE Systems Journal, vol. 12, no. 2, pp. 1414-1422, 2018.
[9] Y. Fujita, K. Shimada, M. Ichihara, and Y. Hamamoto, "A method based on machine learning using hand-crafted features for crack detection from asphalt pavement surface images," in Thirteenth International Conference on Quality Control by Artificial Vision 2017, 2017, vol. 10338, p. 103380I: International Society for Optics and Photonics.
[10] J. H. Seinfeld and S. N. Pandis, Atmospheric chemistry and physics: from air pollution to climate change. John Wiley & Sons, 2016.
[11] Y. Pi, N. D. Nath, and A. H. Behzadan, "Convolutional neural networks for object detection in aerial imagery for disaster response and recovery," Advanced Engineering Informatics, vol. 43, p. 101009, 2020/01/01/ 2020.
[12] B. Mo, R. Li, and J. Dai, "Estimating dynamic origin–destination demand: A hybrid framework using license plate recognition data," Computer-Aided Civil and Infrastructure Engineering, vol. n/a, no. n/a, 2020/01/05 2020.
[13] D. Li, L. Deng, and Z. Cai, "Intelligent vehicle network system and smart city management based on genetic algorithms and image perception," Mechanical Systems and Signal Processing, p. 106623, 2020/01/23/ 2020.
[14] L. Tian, H. Wang, Y. Zhou, and C. Peng, "Video big data in smart city: Background construction and optimization for surveillance video processing," Future Generation Computer Systems, vol. 86, pp. 1371-1382, 2018/09/01/ 2018.
[15] J. A. D. Cameron, P. Savoie, M. E. Kaye, and E. J. Scheme, "Design considerations for the processing system of a CNN-based automated surveillance system," Expert Systems with Applications, vol. 136, pp. 105-114, 2019/12/01/ 2019.
[16] S. Afra and R. Alhajj, "Early warning system: From face recognition by surveillance cameras to social media analysis to detecting suspicious people," Physica A: Statistical Mechanics and its Applications, vol. 540, p. 123151, 2020/02/15/ 2020.
[17] M.-W. Park and I. Brilakis, "Construction worker detection in video frames for initializing vision trackers," Automation in Construction, vol. 28, pp. 15-25, 2012/12/01/ 2012.
[18] C. Chen, Z. Zhu, and A. Hammad, "Automated excavators activity recognition and productivity analysis from construction site surveillance videos," Automation in Construction, vol. 110, p. 103045, 2020/02/01/ 2020.
[19] S. Sufri, F. Dwirahmadi, D. Phung, and S. Rutherford, "A systematic review of Community Engagement (CE) in Disaster Early Warning Systems (EWSs)," Progress in Disaster Science, vol. 5, p. 100058, 2020/01/01/ 2020.
[20] J. Xu, E. Wang, and R. Zhou, "Real-time measuring and warning of surrounding rock dynamic deformation and failure in deep roadway based on machine vision method," Measurement, vol. 149, p. 107028, 2020/01/01/ 2020.
[21] J. Yang, Z. Shi, and Z. Wu, "Vision-based action recognition of construction workers using dense trajectories," Advanced Engineering Informatics, vol. 30, no. 3, pp. 327-336, 2016/08/01/ 2016.
[22] Z. Zhu, X. Ren, and Z. Chen, "Integrated detection and tracking of workforce and equipment from construction jobsite videos," Automation in Construction, vol. 81, pp. 161-171, 2017/09/01/ 2017.
[23] Q. Fang et al., "Detecting non-hardhat-use by a deep learning method from far-field surveillance videos," Automation in Construction, vol. 85, pp. 1-9, 2018/01/01/ 2018.
[24] E. Rezazadeh Azar and B. McCabe, "Part based model and spatial–temporal reasoning to recognize hydraulic excavators in construction images and videos," Automation in Construction, vol. 24, pp. 194-202, 2012/07/01/ 2012.
[25] J. Kim and S. Chi, "Action recognition of earthmoving excavators based on sequential pattern analysis of visual features and operation cycles," Automation in Construction, vol. 104, pp. 255-264, 2019/08/01/ 2019.
[26] R. Akhavian and A. H. Behzadan, "Construction equipment activity recognition for simulation input modeling using mobile sensors and machine learning classifiers," Advanced Engineering Informatics, vol. 29, no. 4, pp. 867-877, 2015/10/01/ 2015.
[27] J. Kim, S. Chi, and J. Seo, "Interaction analysis for vision-based activity identification of earthmoving excavators and dump trucks," Automation in Construction, vol. 87, pp. 297-308, 2018/03/01/ 2018.
[28] J. Gong and H. Caldas Carlos, "Computer Vision-Based Video Interpretation Model for Automated Productivity Analysis of Construction Operations," Journal of Computing in Civil Engineering, vol. 24, no. 3, pp. 252-263, 2010/05/01 2010.
[29] S. M. Silva and C. R. Jung, "Real-Time License Plate Detection and Recognition Using Deep Convolutional Neural Networks," Journal of Visual Communication and Image Representation, p. 102773, 2020/03/04/ 2020.
[30] N. Omar, A. Sengur, and S. G. S. Al-Ali, "Cascaded deep learning-based efficient approach for license plate detection and recognition," Expert Systems with Applications, vol. 149, p. 113280, 2020/07/01/ 2020.
[31] D. Wang, Y. Tian, W. Geng, L. Zhao, and C. Gong, "LPR-Net: Recognizing Chinese license plate in complex environments," Pattern Recognition Letters, vol. 130, pp. 148-156, 2020/02/01/ 2020.
[32] E. J. Sen et al., "Advanced license plate recognition system for car parking," in 2014 International Conference on Embedded Systems (ICES), 2014, pp. 162-165.
[33] G. Hsu, J. Chen, and Y. Chung, "Application-Oriented License Plate Recognition," IEEE Transactions on Vehicular Technology, vol. 62, no. 2, pp. 552-561, 2013.
[34] S. Z. Masood, G. Shu, A. Dehghan, and E. G. J. a. p. a. Ortiz, "License plate detection and recognition using deeply learned convolutional neural networks," 2017.
[35] S. Saraswathi, R. Subban, T. Shanmugasundari, and S. Manogari, "Research on License Plate Recognition Using Digital Image Processing," in 2017 IEEE International Conference on Computational Intelligence and Computing Research (ICCIC), 2017, pp. 1-6.
[36] A. M. Khan, S. M. Awan, M. Arif, Z. Mahmood, and G. Z. Khan, "A Robust Segmentation Free License Plate Recognition Method," in 2019 International Conference on Electrical, Communication, and Computer Engineering (ICECCE), 2019, pp. 1-6.
[37] L. Liu et al., "Deep Learning for Generic Object Detection: A Survey," International Journal of Computer Vision, vol. 128, no. 2, pp. 261-318, 2020/02/01 2020.
[38] W. Puarungroj and N. Boonsirisumpun, "Thai License Plate Recognition Based on Deep Learning," Procedia Computer Science, vol. 135, pp. 214-221, 2018/01/01/ 2018.
[39] Y. Cao, H. Fu, and H. Ma, "An End-to-End Neural Network for Multi-line License Plate Recognition," in 2018 24th International Conference on Pattern Recognition (ICPR), 2018, pp. 3698-3703.
[40] P. Li, M. Nguyen, and W. Q. Yan, "Rotation correction for license plate recognition," in 4th International Conference on Control, Automation and Robotics, ICCAR 2018, April 20, 2018 - April 23, 2018, Auckland, New zealand, 2018, pp. 400-404: Institute of Electrical and Electronics Engineers Inc.
[41] Y. Zhao, Z. Yu, X. Li, and M. Cai, "Chinese license plate image database building methodology for license plate recognition," Journal of Electronic Imaging, vol. 28, no. 1, 2019.
[42] A. Naimi, Y. Kessentini, and M. Hammami, "Multi-nation and Multi-norm License Plates Detection in Real Traffic Surveillance Environment Using Deep Learning," in Neural Information Processing, Cham, 2016, pp. 462-469: Springer International Publishing.
[43] S. Ren, K. He, R. Girshick, and J. Sun, "Faster r-cnn: Towards real-time object detection with region proposal networks," in Advances in neural information processing systems, 2015, pp. 91-99.
[44] H. Li, P. Wang, M. You, and C. Shen, "Reading car license plates using deep neural networks," Image and Vision Computing, vol. 72, pp. 14-23, 2018.
[45] H. Xiang, Y. Zhao, Y. Yuan, G. Zhang, and X. Hu, "Lightweight fully convolutional network for license plate detection," Optik, vol. 178, pp. 1185-1194, 2019.
[46] V. H. Pham, P. Q. Dinh, and V. H. Nguyen, "CNN-Based Character Recognition for License Plate Recognition System," in Intelligent Information and Database Systems, Cham, 2018, pp. 594-603: Springer International Publishing.
[47] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You only look once: Unified, real-time object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 779-788.
[48] Hendry and R.-C. Chen, "Automatic License Plate Recognition via sliding-window darknet-YOLO deep learning," Image and Vision Computing, vol. 87, pp. 47-56, 2019.
[49] J. Redmon and A. Farhadi, "YOLO9000: better, faster, stronger," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 7263-7271.
[50] Y. Jamtsho, P. Riyamongkol, and R. Waranusast, "Real-time Bhutanese license plate localization using YOLO," ICT Express, 2019/11/20/ 2019.
[51] J. Redmon and A. J. a. p. a. Farhadi, "Yolov3: An incremental improvement. arXiv 2018," 8 Apr 2018 2019, Available: https://arxiv.org/abs/1804.02767.
[52] S. Abdullah, M. M. Hasan, and S. M. S. Islam, "YOLO-Based Three-Stage Network for Bangla License Plate Recognition in Dhaka Metropolitan City," in 2018 International Conference on Bangla Speech and Language Processing (ICBSLP), 2018, pp. 1-6.
[53] W. Liu et al., "SSD: Single Shot MultiBox Detector," in Computer Vision – ECCV 2016, Cham, 2016, pp. 21-37: Springer International Publishing.
[54] H. Sun, M. Fu, A. Abdussalam, Z. Huang, S. Sun, and W. Wang, "License Plate Detection and Recognition Based on the YOLO Detector and CRNN-12," in Signal and Information Processing, Networking and Computers, Singapore, 2019, pp. 66-74: Springer Singapore.
[55] R. Laroca et al., "A Robust Real-Time Automatic License Plate Recognition Based on the YOLO Detector," in 2018 International Joint Conference on Neural Networks (IJCNN), 2018, pp. 1-10.
[56] S. M. Silva and C. R. Jung, "Real-Time Brazilian License Plate Detection and Recognition Using Deep Convolutional Neural Networks," in 2017 30th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), 2017, pp. 55-62.
[57] W. Yimyam and M. Ketcham, "The automated parking fee calculation using license plate recognition system," in 2017 International Conference on Digital Arts, Media and Technology (ICDAMT), 2017, pp. 325-329.
[58] L. Xiaoping, L. Feng, L. Xiaoxing, L. Yinxiang, and X. Jianqiang, "Research on character segmentation in license plate recognition," in 4th International Conference on New Trends in Information Science and Service Science, 2010, pp. 345-348.
[59] C. Cheng, L. Mei, and J. Zhang, "License Plate Recognition via Deep Convolutional Neural Network," in 2018 International Conference on Civil and Hydraulic Engineering, IConCHE 2018, November 23, 2018 - November 25, 2018, Qingdao, China, 2018, vol. 189: Institute of Physics Publishing.
[60] P. Gao, Z. Zeng, and S. Sun, "Segmentation-Free Vehicle License Plate Recognition Using CNN," in Signal and Information Processing, Networking and Computers, Singapore, 2019, pp. 50-57: Springer Singapore.
[61] Z. Selmi, M. B. Halima, and A. M. Alimi, "Deep Learning System for Automatic License Plate Detection and Recognition," in 2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), 2017, vol. 01, pp. 1132-1138.
[62] I. N. G. A. Astawa, I. G. N. B. Caturbawa, E. Rudiastari, M. L. Radhitya, and N. K. D. Hariyanti, "Convolutional Neural Network Method Implementation for License Plate Recognition in Android," in 2018 2nd East Indonesia Conference on Computer and Information Technology (EIConCIT), 2018, pp. 176-179.
[63] I. J. Goodfellow, Y. Bulatov, J. Ibarz, S. Arnoud, and V. J. a. p. a. Shet, "Multi-digit number recognition from street view imagery using deep convolutional neural networks," 2013.
[64] T. Nguyen and D. Nguyen, "A New Convolutional Architecture for Vietnamese Car Plate Recognition," in 2018 10th International Conference on Knowledge and Systems Engineering (KSE), 2018, pp. 7-12.
[65] JasonLiTW. (2018/07/27). 實作基於CNN的台鐵訂票驗證碼辨識以及透過模仿及資料增強的訓練集產生器. Available: https://github.com/JasonLiTW/simple-railway-captcha-solver
[66] Z. Song, L. Fu, J. Wu, Z. Liu, R. Li, and Y. Cui, "Kiwifruit detection in field images using Faster R-CNN with VGG16," IFAC-PapersOnLine, vol. 52, no. 30, pp. 76-81, 2019/01/01/ 2019.
[67] J. Riordon, C. McCallum, and D. Sinton, "Deep learning for the classification of human sperm," Computers in Biology and Medicine, vol. 111, p. 103342, 2019/08/01/ 2019.
[68] A. Banan, A. Nasiri, and A. Taheri-Garavand, "Deep learning-based appearance features extraction for automated carp species identification," Aquacultural Engineering, vol. 89, p. 102053, 2020/05/01/ 2020.
[69] F. Chollet, DEEP LEARNING 深度學習必讀Keras大神帶你用Python實作. 旗標科技股份有限公司, 2019.
[70] T.-Y. Lin, P. Dollár, R. Girshick, K. He, B. Hariharan, and S. Belongie, "Feature pyramid networks for object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 2117-2125.
[71] K. Simonyan and A. J. a. p. a. Zisserman, "Very deep convolutional networks for large-scale image recognition," 2014.
[72] S. S. Tabrizi and N. Cavus, "A Hybrid KNN-SVM Model for Iranian License Plate Recognition," Procedia Computer Science, vol. 102, pp. 588-594, 2016/01/01/ 2016.
[73] Y. Kessentini, M. D. Besbes, S. Ammar, and A. Chabbouh, "A two-stage deep neural network for multi-norm license plate detection and recognition," Expert Systems with Applications, vol. 136, pp. 159-170, 2019/12/01/ 2019.
[74] H. Erdinc Kocer and K. Kursat Cevik, "Artificial neural networks based vehicle license plate recognition," Procedia Computer Science, vol. 3, pp. 1033-1037, 2011.
[75] 邱瀅蓁, "河川疏濬工程風險辨識暨知識管理系統設計與開發," 碩士, 營建工程系, 國立臺灣科技大學, 2019.