臺灣博碩士論文加值系統

交通標誌是設立於馬路旁或特定地方的牌匾，主要提供訊息或資訊給予道路使用者。交通標誌使用不同的形狀，顏色和符號或字母來提供不同的資訊和指示。在本研究中，我們將使用廣義霍夫轉換來偵測交通標誌的形狀，廣義霍夫轉換針對符合某些條件下的影像形狀進行投票，得到最多票數的影像參數所代表的形狀即為辨識的結果。廣義霍夫轉換的最主要問題是龐大的計算量及記憶體使用量。為了解決以上問題，我們使用參數表來取代傳統廣義霍夫轉換中的 R -Table，參數表儲存了正規化之交通標誌邊界形狀的参数方程式，使用參數表可有效減少廣義霍夫轉換的計算量。此外，為了減少投票結果的不確定性，我們提出了一種使用使用加權評分以減少候選參數的數量的方法。

Traffic signs or roads signs are the signs that are usually placed on roads to give information or instruction to users. Traffic signs with different shapes, colors and symbols or letters are represented by different meaning and instruction. In this work, we would like to propose a method to recognize the shape of the traffic signs by using modified generalized Hough transform. Generalized Hough transform will vote for certain shape that is under certain condition, the highest vote represent the recognition result. The main problems of generalized Hough transform are huge calculation and memory storage. Hence to solve this problem, we propose a method that replaced the traditional R-table with parameter table. A parameter table is a table that stores the information of normalized shape parametric equation for the normalized shape image. By using the parameter table can reduce the generalized Hough transform efficiently. In addition, to reduce the ambiguity of reference points, we proposed a method by reducing the number of ambiguous value for candidate reference point by using the weighted score method.

Abstract i
Acknowledgements ii
Contents ii
1 Introduction 2
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
2.1 Hough Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Application of Hough Transforms . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 Generalized Hough Transform . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.4 Advantages and Disadvantages of Generalized Hough Transform . . . . . . . 14
2.5 Application of Generalized Hough Transform . . . . . . . . . . . . . . . . . . 15
3 Proposed Method . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
3.1 Preprocessing: Gray-scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2 Preprocessing: Down-sampling and Filtering . . . . . . . . . . . . . . . . . . 18
3.3 Preprocessing: Edge Detection - Sobel Edge Detector . . . . . . . . . . . . . 19
3.4 Preprocessing: Binarization - Ostu Method . . . . . . . . . . . . . . . . . . . 20
3.5 Modied Generalized Hough Transform . . . . . . . . . . . . . . . . . . . . . 21
3.5.1 Replace the R-Table by the Parameter Table . . . . . . . . . . . . . . 21
3.5.2 Voting Process Based on Weighted Scoring Methods . . . . . . . . . . 23
4 Trac Sign Recognition System . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
4.1 Traffic Sign Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2 Experiment Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5 Conclusions and Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

Bibliography
[1] Jabatan Pengangkutan Jalan Malaysia. http://www:jpj:gov:my/.
[2] Ministry of Transportation and Communcation R.O.C. http://www:motc:gov:tw/ch/index:jsp.
[3] John S DaPonte and Martin D Fox. Enhancement of chest radiographs with gradientoperators. Medical Imaging, IEEE Transactions on, 7(2):109{117, 1988.
[4] Paul VC Hough. Machine analysis of bubble chamber pictures. In International Conference on High Energy Accelerators and Instrumentation, volume 73, 1959.
[5] Lei Xu, Erkki Oja, and Pekka Kultanen. A new curve detection method: randomized hough transform (RHT). Pattern recognition letters, 11(5):331{338, 1990.
[6] Ferhat-taleb Alim, K Messaoudi, S Seddiki, O Kerdjidj, et al. Modied circular hough transform using FPGA. In Microelectronics (ICM), 2012 24th International Conference on, pages 1{4. IEEE, 2012.
[7] Wei Wei, Dong Xiaoquan, and Shen Yuanyuan. Research on a two value generalized Hough transform method of identication. In Computer Science and Network Technology (ICCSNT), 2011 International Conference on, volume 1, pages 278{281. IEEE, 2011.
[8] Richard O Duda and Peter E Hart. Use of the Hough transformation to detect lines and curves in pictures. Communications of the ACM, 15(1):11{15, 1972.
[9] Dana H Ballard. Generalizing the Hough transform to detect arbitrary shapes. Pattern recognition, 13(2):111{122, 1981.
[10] How It Works. Hough transform. http://homepages:inf:ed:ac:uk/rbf/HIPR2/hough:htm.
[11] Chun-Pong Chau and Wan-Chi Siu. Adaptive dual-point Hough transform for object
recognition. Computer Vision and Image Understanding, 96(1):1{16, 2004.
[12] HK Yuen, John Illingworth, and Josef Kittler. Detecting partially occluded ellipses
using the Hough transform. Image and Vision Computing, 7(1):31{37, 1989.
[13] P-K Ser and Wan-Chi Siu. Novel 2-d Hough planes for the detection of ellipses. In Speech, Image Processing and Neural Networks, 1994. Proceedings, ISSIPNN'94., 1994 International Symposium on, pages 527{530. IEEE, 1994.
[14] Raymond KK Yip, Peter KS Tam, and Dennis NK Leung. Modication of Hough transform for circles and ellipses detection using a 2-dimensional array. Pattern Recognition,25(9):1007{1022, 1992.
[15] Pui-Kin Ser and Wan-Chi Siu. A new generalized Hough transform for the detection of irregular objects. Journal of Visual Communication and Image Representation,6(3):256{264, 1995.
[16] Raymond KK Yip, Peter KS Tam, and Dennis NK Leung. Modication of Hough
transform for object recognition using a 2-dimensional array. Pattern Recognition,28(11):1733{1744, 1995.
[17] Sheng-Ching Jeng and Wen-Hsiang Tsai. Scale-and orientation-invariant generalized Hough transform- a new approach. Pattern Recognition, 24(11):1037{1051, 1991.
[18] Teach Term:Grayscale. http://www:techterms:com/definition/grayscale.
[19] Rafel C Gonzalez and Richard E Woods. Digital image processing, 2005.
[20] Stephen J Sangwine and Robin EN Horne. The colour image processing handbook.Springer, 1998.
[21] Nobuyuki Otsu. A threshold selection method from gray-level histograms. Automatica,11(285-296):23{27, 1975.
[22] Ren C Luo, H Potlapalli, and DW Hislop. Natural scene segmentation using fractal based autocorrelation. In Industrial Electronics, Control, Instrumentation, and Automation, 1992. Power Electronics and Motion Control., Proceedings of the 1992 International Conference on, pages 700{705. IEEE, 1992.
[23] Ren C Luo, Harsh Potlapalli, and David W Hislop. Translation and scale invariant landmark recognition using receptive eld neural networks. In Intelligent Robots and Systems, 1992., Proceedings of the 1992 lEEE/RSJ International Conference on, volume 1, pages 527{533. IEEE, 1992.
[24] Ren C Luo and Harsh Potlapalli. Landmark recognition using projection learning for mobile robot navigation. In Neural Networks, 1994. IEEE World Congress on Computa-
tional Intelligence., 1994 IEEE International Conference on, volume 4, pages 2703{2708.IEEE, 1994.
[25] Marie de Saint Blancard. Road sign recognition: A study of vision-based decision making for road environment recognition. In Vision-based vehicle guidance, pages 162{172. Springer, 1992.
[26] Giulia Piccioli, Enrico De Micheli, and Marco Campani. A robust method for road sign detection and recognition. In Computer VisionECCV'94, pages 493{500. Springer,1994.
[27] PRO-E. http://www:iw168:cn/proe/zuiquande-ProE-quxianfangcheng-jiaoniruhezhihuizhigezhonggeyangde-ProE-quxian/.
[28] Sebastian Houben, Johannes Stallkamp, Jan Salmen, Marc Schlipsing, and Christian Igel. Detection of trac signs in real-world images: The German Trac Sign Detection Benchmark. In International Joint Conference on Neural Networks, number 1288, 2013.