交通路標在行車安全上扮演了重要的角色，除了規範駕駛的行為之外，也用於指示該路段路況及需注意事項。因此在行車時，路標所提供給駕駛的資訊是很重要的，然而我們在開車時並無法同時注意所有的路標資訊。因此本篇論文提出了一種「即時路標偵測與提取系統」，希望未來能實現於行車紀錄器上。透過行車記錄器的影像加以處理，達到路標偵測與提取的功能，將偵測出來的路標框出，並將夠大的路標提取出來，放置於行車紀錄器畫面上，供駕駛參考並給予駕駛適當的提醒，以提升行車安全。
本系統針對的目標為紅色與藍色路標，形狀有圓形、三角形、倒三角形、八角形、矩形等。系統架構分成兩大部分，分別為色彩空間截取以及形狀識別。色彩空間採用RGB色彩空間，透過此色彩空間將我們感興趣的紅色與藍色區域篩選出來，並透過一些過濾器篩選出符合路標長寬比及大小的物件，再透過分析每個物件內部的顏色分佈情形來判斷其是路標的可能性。而形狀識別部分是使用Distance to Borders(DtB)向量來描述形狀特徵，並運用此特徵搭配Support Vector Machine(SVM)來做形狀分群。若在顏色與形狀分析上都通過篩選，那我們就認定此物件為路標。如此我們便可從整個影像中提取出我們所在意的路標資訊，並將路標資訊顯示於螢幕上供駕駛參考。
從實驗結果可看出。本論文所提出之演算法可以即時的偵測路標，有著高偵測率及適應不同尺寸路標與旋轉角度的特點，甚至在路標被部分遮蔽時也能有不錯的表現。

Road signs play an important role in road safety. In addition to regulating driver behavior, it indicates the road condition and the information that drivers need to pay attention to. Hence, the information provided by road signs is very important; however, we can hardly pay attention to every road sign when driving. Therefore, this study proposes a “real time road sign detection and extraction system” which is used on the event data recorder in the future. The system process the sequential images captured by the event data recorder to detect and extract the information from road signs for the purpose of enhancing driving safety and awareness. The system first detects the road signs, then extracts them and shows them on the screen.
We aim at road signs that are in blue or red color with circular, triangular, octangular or rectangular shapes. The system architecture is composed of color space extraction and shape recognition. This study first uses color information in RGB color space discriminant the red and the blue area. Second, filters distinguish the objects that fit with the road signs’ aspect ratio and size. Third, the color filters analyze the color distribution of every object to determine whether the object is a road sign or not. As to shapes, the study adopts distance to borders (DtB) to depict features of the shapes, and uses the support vector machine (SVM) for shape clustering. If the color and the shape of the object pass through the system filters, the system will predicate it is a road sign. In the end, the system extracts the information from the image and shows it on the screen. From the results, we can conclude that the algorithm can real-time detect road signs and invariant to rotation and scale, even to partial occlusions.

[1]. S. Maldonado-Bascón, S. Lafuente-Arroyo, P. Gil-Jiménez, H. Gómez-Moreno and F. López-Ferreras, “Road-sign detection and recognition based on support vector machines,” IEEE Transactions on Intelligent Transportation Systems. vol. 8, no. 2, pp. 264–278, Jun. 2007
[2]. C. Fang, S. Chen, and C. Fuh, “Road sign detection and tracking,” IEEE Transactions on Vehicular Technology. vol. 52, no. 5, pp. 1329–1341, Sep. 2003.
[3]. H. Liu, D. Liu, and J. Xin, “Real-time recognition of road traffic sign in motion image based on genetic algorithm,” International Conference on Machine Learning and Cybernetics, vol. 1, pp. 83–86, Nov. 2002.
[4]. M. Benallal, J. Meunier, “Real-time color segmentation of road signs,” IEEE Canadian Conference on Electrical and Computer Engineering. vol. 3, pp. 1823-1826. May. 2003.
[5]. L. Estevez and N. Kehtarnavaz, “A real-time histographic approach to road sign recognition,” IEEE Southwest Symposium on Image Analysis and Interpretation, pp. 95–100, Apr. 1996.
[6]. S. Varun, S. Singh, R. S. Kunte, R. D. S. Samuel, and B. Philip, “A road traffic signal recognition system based on template matching employing tree classifier,” Proceedings of the International Conference on Computational Intelligence and Multimedia Applications. pp. 360–365. 2007.
[7]. S. Lafuente-Arroyo, P. Gil-Jiménez, R. Maldonado-Bascón , F. López-Ferreras and S. Maldonado-Bascón, “Traffic sign shape classification evaluation I: SVM using Distance to Borders,” IEEE Intelligent Vehicles Symposium. pp. 557-562, Jun. 2005.
[8]. C.G. Kiran, L.V. Prabhu, V.A. Rahiman, K. Rajeev, A. Sreekumar, “Support vector machine learning based traffic sign detection and shape classification using distance to borders and distance from center features,” TENCON 2008. IEEE Region 10 Conference, pp. 1-6, Nov. 2008.
[9]. C. Cortes, V. Vapnik, “Support-vector networks,” Machine Learning, vol. 20, pp. 273-297, 1995.
[10]. A. Hechri, A. Mtibaa, “Automatic detection and recognition of road sign for driver assistance system,” IEEE Mediterranean Electrotechnical Conference. pp. 888-891, Mar. 2012.
[11]. K. Brkic, "An overview of traffic sign detection methods," Department of Electronics, Microelectronics, Computer and Intelligent Systems Faculty of Electrical Engineering and Computing Unska, vol. 3, 2010.
[12]. H. Huang, C. Chen, Y. Jia, S. Tang, “Automatic detection and recognition of circular road sign” IEEE International Conference on Mechtronic and Embedded Systems and Applications, pp. 626-630, Oct. 2008.
[13]. F. Qin, B. Fang and H.J. Zhao, “Traffic sign segmentation and recognition in scene images,” Chinese Conference on Pattern Recognition (CCPR), pp. 1-5, Oct. 2010
[14]. C.-C. Chang and C.-J. Lin, LIBSVM: A Library for Support Vector Machines, 2001. Available: https://www.csie.ntu.edu.tw/~cjlin/libsvm/
[15]. L. He, Y. Chao, K. Suzuki and K. Wu, “Fast connected-component labeling,” Pattern Recognition, vol. 42, no. 9, pp. 1977-1987, Sep. 2009.
[16]. L. He, Y. Chao and K. Suzuki, “A run-based two-scan labeling algorithm,” IEEE Transactions on Image Processing, vol. 17, no. 5, pp. 749-756, May. 2008.
[17]. F. Larsson, M. Felsberg, “Using Fourier descriptors and spatial models for traffic sign recognition,” Proceedings of the 17th Scandinavian conference on Image analysis, pp. 238-249, May. 2011.
[18]. M. Enzweiler, B. Froehlich, U. Franke, W. Heiden, “Vision-Based Road Sign Detection,” IEEE 18th International Conference on Intelligent Transportation Systems (ITSC), pp.505-510, Sep.2015