隨著科技的發展與推進，無人載具的服務已經不再是遙不可及的願景，而是各個研究者與企業積極投入發展的領域，其中最為大眾所探討以及關心的議題當屬自動駕駛車輛的行車安全性。在可能會牽涉到使用者人身安危的情況下，車輛的安全性可說是自動駕駛技術中最核心必須確保的領域。
道路邊界偵測通常屬於自駕車輛感知模塊的範疇內，自駕車輛通常透過不同的定位演算法來瞭解自身與環境間的關聯及位置，但當定位算法因為受到環境影響產生較大誤差時，道路邊界偵測就會成為確保車輛安全的最後一層屏障，在車輛不確定自身位置時仍能透過道路邊界的偵測結果以免進入無法行駛的區域。在非城市場景中，道路場景通常並未有完整的指引標示，且道路兩旁常緊鄰著危險區域，例如溝渠等會帶來車輛安全威脅的障礙。這些場景使得原先應用在都市場景的車道偵測方法並沒有辦法得到良好成效。
本文提出了一種獨立於定位算法以外的道路邊界偵測及追蹤方法，適用於非都市中並未完整劃設標線的道路場景，以多感測器融合的原則，使用深度相機以及光達(LiDAR)提取各自感測器資料的路緣特徵並加以融合。同時考慮到不同資料來源的測距不同，提出了一個光達資料補償機制以弭平資料範圍的差距。最後將結果利用卡爾曼濾波器進行邊界曲線追蹤，以確保算法的穩定性及強健性。

With the advancement of technology, unmanned vehicle services are no longer a distant dream but a field that researchers and businesses are actively developing. Among the most discussed and concerning issues for the public is the safety of autonomous vehicles. In situations that could potentially involve personal safety risks, vehicle safety is arguably the most critical area that must be ensured in autonomous driving technology.

Road edge detection typically falls within the perception module of autonomous vehicles. Autonomous vehicles often use various positioning algorithms to understand their relationship with the environment and their location. However, when positioning algorithms generate significant errors due to environmental influences, road edge detection becomes the last line of defense to ensure vehicle safety, allowing the vehicle to avoid entering non-navigable areas even when its exact position is uncertain. In non-urban scenarios, road scenes often lack complete guiding signs, and the sides of the roads are frequently adjacent to dangerous areas, such as ditches, which pose safety threats to vehicles. These scenarios make conventional lane detection methods used in urban settings less effective.

This thesis presents a road edge detection and tracking method that operates independently of positioning algorithms, suitable for rural roads without well-defined markings. Using the principle of multi-sensor fusion, it employs depth cameras and LiDAR to extract and combine road edge features from each sensor. Considering the varying measurement ranges of different data sources, a LiDAR data compensation mechanism is introduced to bridge the gap in data coverage. Finally, the results are processed using a Kalman filter for edge curve tracking to ensure the stability and robustness of the algorithm.

[1] “Driverless taxis are coming to the streets of San Francisco” , https://npr.org/2022/06/03/1102922330/driverless-self-driving-taxis-san-francisco-gm-cruise (accessed Jun. 15, 2024)
[2] “Autonomous Driving System Framework” , https://github.com/autowarefoundation/autoware (accessed Jun. 15, 2024)
[3] P. Biber and W. Strasser, "The Normal Distributions Transform: a New Approach to Laser Scan Matching," in Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003) (Cat. No.03CH37453),2003.
[4] H. Wang and B. Liu, "G-NET: Accurate Lane Detection Model for Autonomous Vehicle," IEEE Systems Journal, vol. 17, no. 2, pp. 2039-2048, June 2023
[5] S. D. Vidya Sagar and C. J. Prabhakar, "A Vison Based Lane Detection Approach Using Vertical Lane Finder Method," in 2022 International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics ( DISCOVER), Shivamogga, India, 2022, pp. 207-211.
[6] R. E. Kalman, “A New Approach to Linear Filtering and Prediction Problems,” Transaction of the ASME - Journal of Basic Engineering, pp. 35-45, 1960.
[7] S. Yin and L. Zhu, "Country Road Boundaries Detection for Autonomous Driving Based on LiDAR Point Cloud," in 2022 IEEE 5th International Conference on Electronics Technology (ICET), Chengdu, China, 2022, pp. 1117-1120.
[8] Martin A. Fischler and Robert C. Bolles, “Random Sample Consensus: a Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography”, Communications of the ACM, vol.24, no.6 , pp.381-395, June 1981.
[9] D. Guo, G. Yang, B. Qi and C. Wang, "Curb Detection and Compensation Method for Autonomous Driving via a 3-D-LiDAR Sensor," IEEE Sensors Journal, vol. 22, no. 20, pp. 19500-19512, October 2022.
[10] N. Sukumar and P. Sumathi, "A Robust Vision-based Lane Detection using RANSAC Algorithm," in 2022 IEEE Global Conference on Computing, Power and Communication Technologies (GlobConPT), New Delhi, India, 2022, pp. 1-5.
[11] H. Komori and K. Onoguchi, "Driving Lane Detection Based on Recognition of Road Boundary Situation," in 2018 Digital Image Computing: Techniques and Applications (DICTA), Canberra, ACT, Australia, 2018, pp. 1-8.
[12] K. Li, H. Xiong, D. Yu, J. Liu, Y. Guo and J. Wang, "An End-to-End Multi-Task Learning Model for Drivable Road Detection via Edge Refinement and Geometric Deformation," IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 7, pp. 8641-8651, July 2022.
[13] D. Patel and K. Elgazzar, "Road Boundary Detection using Camera and mmwave Radar," in 2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA), Cairo, Egypt, 2022, pp. 1-7.
[14] M. A. Rasyidy, Y. Y. Nazaruddin and A. Widyotriatmo, "Regression with Ensemble of RANSAC in Camera-LiDAR Fusion for Road Boundary Detection and Modeling," in 2022 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), Bedford, United Kingdom, 2022, pp. 1-6.
[15] E. Debie et al., "Multimodal Fusion for Objective Assessment of Cognitive Workload: A Review," IEEE Transactions on Cybernetics, vol. 51, no. 3, pp. 1542-1555, March 2021.
[16] Ya-Jin Hsiao, “Real-time Golf Course Road Boundary Detection and Tracking for Autonomous Vehicles”, M.S. thesis, National Cheng Kung University, 2023.
[17] Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff and Hartwig Adam, “Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation”, in European Conference on Computer Vision, 2018.
[18] K.K.Ayten and O.Kurnaz, “Real-Time Implementation of Image Based PLC Control for a Robotic Platform”, Balkan Journal of Electrical and Computer Engineering, vol.7, No.1, pp. 64-71, January 2019.
[19] M. Himmelsbach, F. v. Hundelshausen and H . -J. Wuensche, "Fast Segmentation of 3D point clouds for ground vehicles," in 2010 IEEE Intelligent Vehicles Symposium, La Jolla, CA, USA, 2010, pp. 560-565.
[20] J. H. Song, S. H. Han, K. Y. Yu and Y. I. Kim, "Assessing the Possibility of Land-cover Classification using LiDAR Intensity Data", International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. 34, pp. 259-262, 2002.
[21] G. Neuhold, T. Ollmann, S. R. Bulò and P. Kontschieder, "The Mapillary Vistas Dataset for Semantic Understanding of Street Scenes,"in 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 2017, pp. 5000-5009.
[22] A. Geiger, P. Lenz and R. Urtasun, "Are we ready for autonomous driving? The KITTI vision benchmark suite," in 2012 IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA, 2012, pp. 3354-3361.
[23] J. Canny, "A Computational Approach to Edge Detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. PAMI-8, no. 6, pp. 679-698, Nov. 1986.
[24] “E-Z-GO Freedom RXV golf cart”, https://ezgo.txtsv.com/sites/default/files/2020-10/Freedom%20RXV.pdf (accessed Jun. 15, 2024)
[25] “Robosense LiDAR M1”, https://www.robosense.ai/en/rslidar/RS-LiDAR-M1 (accessed Jun. 25, 2024)
[26] “Intel RealSense Depth Camera D455”, https://www.intelrealsense.com/depth-camera -d455 (accessed Jun. 25, 2024)
[27] “u-blox C100-F9K application board”, https://www.u-blox.com/en/product/c100-f9k-application-board (accessed Jun. 25, 2024)
[28] “Neousys Technology: Nuvo-5095GC”, https://www.neousys-tech.com/en/product/ product-lines/edge-ai-gpu-computing/nuvo-5095gc-gpu-computer (accessed Jun. 25, 2024)
[29] “Robot Operating System”, https://www.ros.org/ (accessed Jun. 25, 2024)
[30] F. Poggenhans et al., "Lanelet2: A high-definition map framework for the future of automated driving," in 2018 21st International Conference on Intelligent Transportation Systems (ITSC), Maui, HI, USA, 2018, pp. 1672-1679.
[31] “Vector Map Builder”, https://tools.tier4.jp/vector_map_builder_ll2/ (accessed Jul. 10, 2024)
[32] “Autoware”, https://autoware.org/ (accessed Jul. 10, 2024)
[33] Q. Zou, H. Jiang, Q. Dai, Y. Yue, L. Chen and Q. Wang, "Robust Lane Detection From Continuous Driving Scenes Using Deep Neural Networks," IEEE Transactions on Vehicular Technology, vol. 69, no. 1, pp. 41-54, Jan. 2020.