臺灣博碩士論文加值系統

科技執法是現今改善交通安全的重要方法之一，因交通事故件數與事故率逐年攀升，輕則財損，重則人亡，而事故發生原因皆起於交通違規，加上台灣高齡化與少子化社會問題遲遲無法得到改善，導致警察人力嚴重缺乏，科技執法在未來更是不可或缺的重要議題。運用人工智慧進行優化民眾用路安全，減少交通上的違規行為，從而減少事故和擁堵的發生。
本研究提出了一種惡劣氣候影像改善並結合路口人工智慧科技執法之路側系統，以解決交通管理中的交通事故和擁堵等重要問題。 為了改善上述之問題，此系統運用YOLOv7技術準確辨識車輛，結合DeepSORT追蹤車流動向並搭配OpenCV規劃車道違規區域進行科技執法應用。借助這些技術，智慧路側系統可以更靈敏的檢測道路上的車輛行為，更好的協調交通，提高道路通行效率。總體而言，這種智慧交通號誌燈的智慧路側系統有望在城市交通管理中發揮重要作用，有效減少交通事故和擁堵的發生，提高道路通行效率，為民眾提供更安全、更便捷的出行。

Technological law enforcement is one of the important methods to improve traffic safety today. With the increasing number and rate of traffic accidents, ranging from minor property damage to severe casualties, the causes of accidents are mostly related to traffic violations. Coupled with the aging population and low birth rate issues in Taiwan, which have yet to be effectively addressed, there is a severe shortage of police manpower. Therefore, the use of technology in law enforcement is an indispensable and crucial topic for the future. By applying artificial intelligence (AI) to optimize public road safety and reduce traffic violations, the occurrence of accidents and congestion can be minimized.
This study proposes a roadside system that combines the improvement of adverse weather image recognition with AI technology for intersection law enforcement, aiming to address important issues in traffic management, such as traffic accidents and congestion. To improve the aforementioned problems, the system utilizes YOLOv7 technology to accurately identify vehicles, combines DeepSORT for tracking traffic flow, and utilizes OpenCV to define areas for enforcing technological law enforcement in lane violations. With the assistance of these technologies, the intelligent roadside system can detect vehicle behavior on the road more sensitively, better coordinate traffic, and improve road traffic efficiency. Overall, this intelligent roadside system with intelligent traffic signal lights is expected to play a significant role in urban traffic management, effectively reducing the occurrence of traffic accidents and congestion, and enhancing road traffic efficiency to provide the public with safer and more convenient travel options.

摘要.....i
Abstract......ii
誌謝.....iii
目錄.....iv
表目錄.....vi
圖目錄.....vii
第一章 緒論.....1
1.1 研究背景與動機.....1
1.2 研究目的與方法.....3
1.3 文獻探討.....4
1.4 論文組織架構.....12
第二章 相關技術探討.....12
2.1 AOD–Net.....12
2.2 YOLOv7.....14
2.3 多目標追蹤檢測.....16
2.3.1 SORT.....16
2.3.2 DeepSORT.....17
2.3 YOLOv7-DeepSORT.....19
2.4車牌辨識.....21
第三章 研究方法與流程.....22
3.1 系統架構.....23
3.1.1 偵測.....23
3.1.2 追蹤.....24
3.1.3 預測.....24
3.2 YOLOv7-DeepSORT目標ID+追蹤.....25
3.3 資料來源.....26
第四章 系統整合與實作結果.....27
4.1 實驗環境.....27
4.2 實驗設備.....28
4.3 實驗資料集.....29
4.4 實驗方法.....29
4.5 實驗結果.....30
4.5.1紅燈壓線.....33
4.5.2 違規變換車道.....35
4.5.3 未禮讓行人.....37
第五章 結論與未來展望.....40
參考文獻.....41
Extended Abstract.....44

[1]B. Cai, X. Xu, K. Jia, C. Qing, and D. J. I. t. o. i. p. Tao, "Dehazenet: An end-to-end system for single image haze removal," vol. 25, no. 11, pp. 5187-5198, 2016.
[2]L. F. P. De Oliveira, L. T. Manera, and P. D. G. J. I. I. o. T. J. Da Luz, "Development of a smart traffic light control system with real-time monitoring," vol. 8, no. 5, pp. 3384-3393, 2020.
[3]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: IEEE.
[4]R. J. Franklin, "Traffic signal violation detection using artificial intelligence and deep learning," in 2020 5th international conference on communication and electronics systems (ICCES), 2020, pp. 839-844: IEEE.
[5]M. A. B. Zuraimi and F. H. K. Zaman, "Vehicle detection and tracking using YOLO and DeepSORT," in 2021 IEEE 11th IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE), 2021, pp. 23-29: IEEE.
[6]K.-H. N. Bui, H. Yi, H. Jung, and J. Cho, "Video-based traffic flow analysis for turning volume estimation at signalized intersections," in Asian Conference on Intelligent Information and Database Systems, 2020, pp. 152-162: Springer.
[7]B. Li, X. Peng, Z. Wang, J. Xu, and D. Feng, "Aod-net: All-in-one dehazing network," in Proceedings of the IEEE international conference on computer vision, 2017, pp. 4770-4778.
[8]R. J. A. i. n. i. p. s. Faster, "Towards real-time object detection with region proposal networks," vol. 9199, no. 10.5555, pp. 2969239-2969250, 2015.
[9]K. He, J. Sun, X. J. I. t. o. p. a. Tang, and m. intelligence, "Single image haze removal using dark channel prior," vol. 33, no. 12, pp. 2341-2353, 2010.
[10]W. Ren, S. Liu, H. Zhang, J. Pan, X. Cao, and M.-H. Yang, "Single image dehazing via multi-scale convolutional neural networks," in Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part II 14, 2016, pp. 154-169: Springer.
[11]E. J. J. N. Y. McCartney, "Optics of the atmosphere: scattering by molecules and particles," 1976.
[12]S. G. Narasimhan and S. K. Nayar, "Chromatic framework for vision in bad weather," in Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No. PR00662), 2000, vol. 1, pp. 598-605: IEEE.
[13]S. G. Narasimhan and S. K. J. I. j. o. c. v. Nayar, "Vision and the atmosphere," vol. 48, pp. 233-254, 2002.
[14]M. Sulami, I. Glatzer, R. Fattal, and M. Werman, "Automatic recovery of the atmospheric light in hazy images," in 2014 IEEE International Conference on Computational Photography (ICCP), 2014, pp. 1-11: IEEE.
[15]G. Meng, Y. Wang, J. Duan, S. Xiang, and C. Pan, "Efficient image dehazing with boundary constraint and contextual regularization," in Proceedings of the IEEE international conference on computer vision, 2013, pp. 617-624.
[16]J.-P. Tarel and N. Hautiere, "Fast visibility restoration from a single color or gray level image," in 2009 IEEE 12th international conference on computer vision, 2009, pp. 2201-2208: IEEE.
[17]D. Berman and S. Avidan, "Non-local image dehazing," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 1674-1682.
[18]Q. Zhu, J. Mai, and L. J. I. t. o. i. p. Shao, "A fast single image haze removal algorithm using color attenuation prior," vol. 24, no. 11, pp. 3522-3533, 2015.
[19]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," in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 7464-7475.
[20]W. Luo, J. Xing, A. Milan, X. Zhang, W. Liu, and T.-K. J. A. i. Kim, "Multiple object tracking: A literature review," vol. 293, p. 103448, 2021.
[21]Z. Wang, L. Zheng, Y. Liu, Y. Li, and S. Wang, "Towards real-time multi-object tracking," in European Conference on Computer Vision, 2020, pp. 107-122: Springer.
[22]N. Wojke, A. Bewley, and D. Paulus, "Simple online and realtime tracking with a deep association metric," in 2017 IEEE international conference on image processing (ICIP), 2017, pp. 3645-3649: IEEE.
[23]A. Bewley, Z. Ge, L. Ott, F. Ramos, and B. Upcroft, "Simple online and realtime tracking," in 2016 IEEE international conference on image processing (ICIP), 2016, pp. 3464-3468: IEEE.
[24]A. Alahi, P. Vandergheynst, M. Bierlaire, M. J. C. V. Kunt, and I. Understanding, "Cascade of descriptors to detect and track objects across any network of cameras," vol. 114, no. 6, pp. 624-640, 2010.
[25]E. Brookner, "Tracking and Kalman filtering made easy," 1998.
[26]E. Hamuda, B. Mc Ginley, M. Glavin, E. J. C. Jones, and e. i. agriculture, "Improved image processing-based crop detection using Kalman filtering and the Hungarian algorithm," vol. 148, pp. 37-44, 2018.
[27]Q. Wang, "License plate recognition via convolutional neural networks," in 2017 8th IEEE International Conference on Software Engineering and Service Science (ICSESS), 2017, pp. 926-929: IEEE.
[28]T.-Y. Lin et al., "Microsoft coco: Common objects in context," in Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13, 2014, pp. 740-755: Springer.
[29]F. Ciaglia, F. S. Zuppichini, P. Guerrie, M. McQuade, and J. J. a. p. a. Solawetz, "Roboflow 100: A Rich, Multi-Domain Object Detection Benchmark," 2022.
[30]A. Mohamed, K. Qian, M. Elhoseiny, and C. Claudel, "Social-stgcnn: A social spatio-temporal graph convolutional neural network for human trajectory prediction," in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 14424-14432.
[31]Y. Ma, X. Zhu, S. Zhang, R. Yang, W. Wang, and D. Manocha, "Trafficpredict: Trajectory prediction for heterogeneous traffic-agents," in Proceedings of the AAAI conference on artificial intelligence, 2019, vol. 33, no. 01, pp. 6120-6127.