本論文基於影像定位與自主控制技術，研發一動態路徑規劃與即時避障系統，分為：「機電動力控制與轉向系統」、「影像定位系統」以及「車輛決策系統」三部分來進行討論。「機電動力控制與轉向系統」以改裝市售電動三輪車作為本論文實驗驗證之無人駕駛平台，結合霍爾感測器(Hall Sensor)與PD(Proportional–Derivate)即時控制，使系統具有良好的即時性與準確性；「影像定位系統」則利用攝影機與目標物底部位置之光學幾何關係及深度學習技術，定位標的物及人所在位置；「車輛決策系統」則為使用人工位能場(Artificial Potential Field, APF)方式做及時規劃行駛路徑，並透過三角定位與單標的物方式推算車輛位置及車輛航向角方向角來進行路徑追蹤，使自主導航三輪車得以自動且精準地在各種條件下運作，最後搭配人性化操作介面，使用者能夠在介面中查看車輛位置、標的物位置與障礙物位置以及相關參數。
為了測試定位技術與避障路徑規劃技術之平台，在相同實驗條件下本文多次測試S型閃避標的物以及直線行駛時閃避行人，經過驗證，皆可即時規劃出適合的路徑並準確地完成閃避標的物及人，。故本實驗之自駕平台能夠在行駛狀態下達到自動避障功能。

This study focused on creating one dynamic path planning and real-time obstacle avoidance system. The whole architecture are divided into three parts: “electromechanical control and steering module”, “vehicle locating system with computer vision” and “vehicle decision system”. In the first part, we modified the electric tricycle as the autonomous vehicle for the confirmation purpose. With Hall sensor and Proportional-Derivate control included, the platform can be even more precise and immediate. The second part, using the optical relationship between camera and the bottom of the target cooperate with deep learning technology to locate the precise position of target and pedestrian. The third part of the system utilize Artificial Potential Field to execute the real-time path planning. Also, implement triangular positioning and single target recognition to reckon vehicle position and heading angle. Thus, it can make the autonomous system much precise under various condition. At last, establishing friendly user interface and providing the information and parameters to the current status.
In order to evaluate the accuracy and stability of this system, testing several S-curve avoidances with targets and pedestrian avoidance in straight under the same condition. According to the experiment, this system can plan the suitable and accurate path and then avoid the obstacles perfectly. The result was capable and convinced.

[1]A. Alam and Z. Jaffery, “A computer vision system for detection and avoidance for automotive vehicles,” Annual IEEE India Conference (INDICON), pp. 1–6, 2015.
[2]Y. Huang and S. Liu, “Multi-class obstacle detection and classification using stereovision and improved active contour models,” IET Intelligent Transport Systems, vol. 10, issue. 3, pp. 197–205, 2016.
[3]V. D. Nguyen, H. V. Nguyen, D. T. Tran, S. J. Lee and J. W. Jeon, “Learning Framework for Robust Obstacle Detection, Recognition, and Tracking,” IEEE Transactions on Intelligent Transportation Systems, vol. 18, issue. 7, pp. 1633–1646, 2017.
[4]Y. Tsai, K. Chen, Y. Chen and J. Cheng, “Accurate and fast obstacle detection method for automotive applications based on stereo vision,” International Symposium on VLSI Design, Automation and Test (VLSI-DAT), pp.1–4, 2018.
[5]O. Kaiwartya, Y. Cao, J. Lloret, S. Kumar, N. Aslam, R. Kharel, A. H. Abdullah, R. R. Shah, “Geometry-based Localization for GPS Outage in Vehicular Cyber Physical Systems,” IEEE Transactions on Vehicular Technology, vol. 67, issue. 5, pp. 3800–3812, 2018.
[6]K. Sundar, S. Misra and S. Rathinam, “Routing unmanned vehicles in GPS-denied environments,” IEEE International Conference on International Conference on Unmanned Aircraft Systems, pp. 4579–4591, 2017.
[7]D. Cui, J. Xue and N. Zheng, “Real-Time Global Localization of Robotic Cars in Lane Level via Lane Marking Detection and Shape Registration,” IEEE Transactions on Intelligent Transportation Systems, vol. 17, issue. 4, pp. 1039–1050, 2016.
[8]K. Jo, Y. Jo and J. K. Suhr, “Precise Localization of an Autonomous Car Based on Probabilistic Noise Models of Road Surface Marker Features Using Multiple Cameras,” IEEE Transactions on Intelligent Transportation Systems, vol. 16, issue. 6, pp. 3377–3392, 2015.
[9]Y. Gu, Y. Wada, L. Hsu and S. Kamijo, “Vehicle self-localization in urban canyon using 3D map based GPS positioning and vehicle sensors,” IEEE Conferences on Connected Vehicles and Expo , pp. 792–798, 2014.
[10]C. Kuo, H. Chou, S. Chi and Y. Lien, “Vision-Based Obstacle Avoidance Navigation with Autonomous Humanoid Robots for Structured Competition Problems,” IJHR International Journal of Humanoid Robotics, vol. 10, no. 3, pp. 1–36, 2013.
[11]J. Park, JY. Kim, BC. Kim and S. Kim ,“Global Map Generation using LiDAR and Stereo Camera for Initial Positioning of Mobile Robot,” International Conference on Information and Communication Technology Robotics (ICT-ROBOT), pp. 1–4,2018.
[12]I. Chabini and S. Lan, “Adaptations of the A* algorithm for the Computation of Fastest Paths in Deterministic Discrete-Time Dynamic Networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 3, no. 1, pp. 60–74, 2002.
[13]張傑，「以改良的A*演算法規劃較佳導引路徑之研究」，碩士學位論文，大同大學，民國98年。
[14]張書豪，「具自主導航與避障之嵌入式移動機器人控制系統開發」，碩士學位論文，國立臺灣科技大學，民國99年。
[15]W. Danwei and Q. Feng, “Trajectory planning for a four-wheel-steering vehicle,” in Robotics and Automation, 2001. Proceedings 2001 ICRA. IEEE International Conference on, 2001, pp. 3320-3325 vol.4.
[16]X. Guangyan and W. Danwei, “Full State Tracking of a Four-Wheel-Steering Vehicle based on Output Tracking Control Strategies, ” in Control, Automation, Robotics and Vision, 2006. ICARCV '06. 9th International Conference on, 2006, pp. 1-6.
[17]A. Cherubini, F. Spindler, and F. Chaumette, “Autonomous visual navigation and laser-based moving obstacle avoidance,” IEEE Transactions on Intelligent Transportation Systems, vol. 15, issue. 5, pp. 2101–2110, 2014.
[18]J. Lee, B.Y. Kang, and D.W. Kim, “Fast genetic algorithm for robot path planning,” Electronics Letters, vol. 49, pp. 1449–1451, 2013.
[19]J.A. Oroko and G.N. Nyakoe, “Obstacle avoidance and path planning schemes for autonomous navigation of a mobile robot: A Review”, Proceedings of the 2012 Mechanical Engineering Conference on Sustainable Research and Innovation, vol. 4, pp. 314-318, 2012.
[20]D. Cong, C. Liang, Q. Gong, X. Yang and J. Liu, “Path Planning and Following of Omnidirectional Mobile Robot Based on B-spline,” Chinese Control And Decision Conference (CCDC), pp. 4931–4936, 2018.
[21]E. Elsheikh, M. El-Bardini and M. Fkirin “Practical path planning and path following for a non-holonomic mobile robot based on visual servoing,” IEEE Information Technology, Networking, Electronic and Automation Control Conference, pp.401–406, 2016.
[22]J. Redmon and A. Farhadi, ‘‘YOLOv3: An incremental improvement,’’ 2018, arXiv:1804.02767. [Online]. Available: https://arxiv.org/abs/1804.02767