臺灣博碩士論文加值系統

隨著現代社會的發展，汽車已成為人們日常生活中不可或缺的交通工具。然而隨著汽車的普及交通事故也隨之增加，尤其是由視線死角引起的事故屢見不鮮。視線死角是指駕駛者無法清楚看到的區域，這可能導致駕駛者對交通狀況不清楚或反應不及，進而引發意外事故。根據研究，傳統的視線死角解決方法如後視鏡和感測器都存在局限性，無法完全消除風險。隨著自動駕駛技術的興起，如何在駕駛過程中進一步提升駕駛員的感知能力和反應時間成為了研究的重點。現有的駕駛輔助系統（ADAS）在增強駕駛員的意識和反應時間方面取得了顯著進展，但在解決視線死角問題上仍然不足。深度學習和影像處理技術為解決此問題提供了新的方法。這些技術能夠透過數據訓練模型，實現更加精確和高效的影像識別和處理，從而為駕駛者提供更全面的道路狀況。此外這些技術還可以與其他技術如毫米波雷達相互結合，進一步提高系統的準確性和可靠性。這不僅有助於改善現有的駕駛輔助系統還可推動無人駕駛技術的發展最終達成更智慧的交通環境。為解決這一問題，本研究提出了一種基於深度學習和影像處理技術的解決方案。該方案利用影像傳輸技術將前方車輛的視角傳輸給本車輛，並透過影像縫合技術融合兩個視角的特定影像，使本車輛能夠看到前方車輛所遮擋的區域，從而避免視線死角對交通安全的影響。這技術提供了更詳盡的交通資訊，來降低了交通事故的風險。本研究透過搭建自走車模型與環境，進行了一系列實驗來驗證所提方案的可行性和有效性。

With the rise of modern society, cars have become essential in daily life. However, the increase in car use has led to more frequent traffic accidents, especially those caused by blind spots. Blind spots are Areas that drivers cannot see, potentially causing unclear traffic conditions and delayed reactions, leading to accidents. Traditional solutions like rearview mirrors and sensors have limitations and cannot fully eliminate risks. Therefore, there is an urgent need for advanced technologies to improve vehicle visibility and enhance road safety. Existing advanced driver assistance systems (ADAS) have made progress but still fall short in addressing blind spot issues. Recent advancements in deep learning and image processing offer new opportunities by training models on large-scale data for accurate image recognition and processing, providing comprehensive road condition information.
This study proposes a solution using deep learning and image processing technologies. It transmits the perspective of the vehicle ahead and merges images through image stitching, allowing the current vehicle to see Areas blocked by the vehicle in front, thus reducing the impact of blind spots on traffic safety. This technology provides detailed traffic information, enhances drivers' perception, and improves reaction capabilities, reducing accident risks. Experiments with an autonomous vehicle model verify the solution's feasibility and effectiveness. Results show the system enhances drivers' visual range and reaction time, effectively reducing traffic accident risks from blind spots. Future applications could further improve traffic system intelligence and safety.

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...1
1.2研究動機與目的...1
1.3章節安排...2
第二章 文獻回顧...3
2.1 車聯網...3
2.2 Socket傳輸技術概述...4
2.2.1 TCP 協定...4
2.2.2 UDP協定...5
2.3 深度學習(You Only Look Once, YOLO)...6
2.3.1 YOLOv9...9
2.4 JPEG圖像壓縮技術...9
2.5 車輛控制方法...13
2.5.1模糊控制器...14
2.5.2比例-積分-微分控制器...15
第三章 實驗方法...17
3.1實驗設備...17
3.2 系統架構...20
3.2.1 硬體架構...20
3.2.2 軟體架構...21
3.3 系統流程...22
3.3.1 前車系統...23
3.3.2 本車系統(PC端程式執行流程圖)...25
3.3.3縫合透視系統...26
3.3.4 車輛控制系統...28
3.3.5 顯示系統...29
3.4 實驗方法...30
3.4.1 場地布置...30
3.4.2 系統啟動進行影像傳輸...32
3.4.3 YOLO物體識別...35
3.4.4 影像縫合透視...36
3.4.5 馬達控制...38
3.4.6 視窗顯示...41
第四章 實驗與結果...43
4.1影像壓縮比例比較...43
4.2 場地辨識與分割結果...47
4.3縫合透視功能效果...48
4.4情境功能測試比較...50
第五章 結論與未來展望...53
參考文獻...54
Extended Abstract...57

 

[1]World Health Organization (WHO). (2023). Global status report on road safety 2023. Retrieved from https://www.who.int/
[2]Console and Associates. (2024). Blind spot car accident injuries. Retrieved from https://natlawreview.com/
[3]More, S., Mulla, A. C., Argade, S. G., Raskar, M., Sakhare, P., & Jadhav, S. P. (2024). Advanced Driver Assistance Systems (ADAS) feature in Modern Autonomous Vehicle. 2024 3rd International Conference for Innovation in Technology (INOCON), 1-5. https://doi.org/10.1109/INOCON60754.2024.10511531
[4]Lin, B.-F., et al. (2012). Integrating appearance and edge features for sedan vehicle detection in the blind-spot area. IEEE Transactions on Intelligent Transportation Systems, 13(2), 737-747. https://doi.org/10.1109/TITS.2011.2182649
[5]Boxu, Z., Cheng, Z., Libin, Z., & Meixia, L. (2023). Optimization of vehicle blind zone monitoring (BSD) evaluation scheme based on image processing and radar ranging algorithm. 2023 IEEE 2nd International Conference on Electrical Engineering, Big Data and Algorithms (EEBDA), 1657-1662. https://doi.org/10.1109/EEBDA56825.2023.10090708
[6]Nilsson, J., Ödblom, A. C. E., Fredriksson, J., Zafar, A., & Ahmed, F. (2010). Performance evaluation method for mobile computer vision systems using augmented reality. 2010 IEEE Virtual Reality Conference (VR), 19-22. https://doi.org/10.1109/VR.2010.5444821
[7]PJ Campanaro Attorney at Law. (2023). Types of car accidents caused by blind spots. Retrieved from https://www.csralawyer.com/.
[8]Yuan, B., Chen, Y.-A., & Ye, S. (2018). A lightweight augmented reality system to see-through cars. 2018 7th International Congress on Advanced Applied Informatics (IIAI-AAI), 855-860. https://doi.org/10.1109/IIAI-AAI.2018.00174
[9]Rameau, F., Ha, H., Joo, K., Choi, J., Park, K., & Kweon, I. S. (2016). A real-time augmented reality system to see-through cars. IEEE Transactions on Visualization and Computer Graphics, 22(11), 2395-2404. https://doi.org/10.1109/TVCG.2016.2593768
[10]Park, B.-J., Yoon, C., Lee, J.-W., & Kim, K.-H. (2015). Augmented reality based on driving situation awareness in vehicle. 2015 17th International Conference on Advanced Communication Technology (ICACT), 593-595. https://doi.org/10.1109/ICACT.2015.7224865
[11]Gomes, P., Olaverri-Monreal, C., & Ferreira, M. (2012). Making vehicles transparent through V2V video streaming. IEEE Transactions on Intelligent Transportation Systems, 13(2), 930-938. https://doi.org/10.1109/TITS.2012.2188289
[12]Li, H., & Nashashibi, F. (2011). Multi-vehicle cooperative perception and augmented reality for driver assistance: A possibility to ‘see’ through front vehicle. 2011 14th International IEEE Conference on Intelligent Transportation Systems (ITSC), 242-247. https://doi.org/10.1109/ITSC.2011.6083061
[13]Borsoi, R. A., & Costa, G. H. (2018). On the performance and implementation of parallax free video see-through displays. IEEE Transactions on Visualization and Computer Graphics, 24(6), 2011-2022. https://doi.org/10.1109/TVCG.2017.2705184
[14]Olaverri-Monreal, C., Gomes, P., Fernandes, R., Vieira, F., & Ferreira, M. (2010). The See-Through System: A VANET-enabled assistant for overtaking maneuvers. 2010 IEEE Intelligent Vehicles Symposium, 123-128. https://doi.org/10.1109/IVS.2010.5548020
[15]Gomes, P., Vieira, F., & Ferreira, M. (2012). The See-Through System: From implementation to test-drive. 2012 IEEE Vehicular Networking Conference (VNC), 40-47. https://doi.org/10.1109/VNC.2012.6407443
[16]Wang, Z., Jin, Q., & Wu, B. (2022). Design of a vision blind spot detection system based on depth camera. 2022 IEEE International Conference on Dependable, Autonomic and Secure Computing, International Conference on Pervasive Intelligence and Computing, International Conference on Cloud and Big Data Computing, International Conference on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech), 1-5. https://doi.org/10.1109/DASC/PiCom/CBDCom/Cy55231.2022.9927963
[17]Xia, Z., Gong, J., Long, Y., Ren, W., Wang, J., & Lan, H. (2022). Research on traffic accident detection based on vehicle perspective. 2022 4th International Conference on Robotics and Computer Vision (ICRCV), 223-227. https://doi.org/10.1109/ICRCV55858.2022.9953179
[18]Park, B.-J., Yoon, C., Lee, J.-W., & Kim, K.-H. (2015). Augmented reality based on driving situation awareness in vehicle. 2015 17th International Conference on Advanced Communication Technology (ICACT), 593-595. https://doi.org/10.1109/ICACT.2015.7224865
[19]https://www.cw.com.tw/article/5122137
[20]https://www.geeksforgeeks.org/differences-between-tcp-and-udp/
[21]Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 779-788. https://doi.org/10.1109/CVPR.2016.91
[22]Bakirci, M., & Bayraktar, I. (2024). YOLOv9-enabled vehicle detection for urban security and forensics applications. 2024 12th International Symposium on Digital Forensics and Security (ISDFS), 1-6. https://doi.org/10.1109/ISDFS60797.2024.10527304
[23]Nusari, A. N. M., Ozbek, I. Y., & Oral, E. A. (2024). Automatic vehicle accident detection and classification from images: A comparison of YOLOv9 and YOLO-NAS algorithms. 2024 32nd Signal Processing and Communications Applications Conference (SIU), 1-4. https://doi.org/10.1109/SIU61531.2024.10600761
[24]Shen, Y., & Yan, W. Q. (2018). Blind spot monitoring using deep learning. 2018 International Conference on Image and Vision Computing New Zealand (IVCNZ), 1-5. https://doi.org/10.1109/IVCNZ.2018.8634716
[25]Good, W. F., Maitz, G. S., & Gur, D. (1994). Joint photographic experts group (JPEG) compatible data compression of mammograms. Journal of Digital Imaging, 7(3), 123-132.
[26]夏漢軒（2022）。基於ROS2.0分散式通訊架構結合無線隨意網路之多路側單元自動駕駛系統研究。﹝碩士論文。國立虎尾科技大學﹞臺灣博碩士論文知識加值系統。 https://hdl.handle.net/11296/nsj747。
[27]Chandan, G., Jain, A., Jain, H., & Mohana. (2018). Real time object detection and tracking using deep learning and OpenCV. 2018 International Conference on Inventive Research in Computing Applications (ICIRCA), 1305-1308. https://doi.org/10.1109/ICIRCA.2018.8597266