臺灣博碩士論文加值系統

由於開發中國家發展迅速，車輛的需求大量成長，因此交通管理日趨繁雜，而車牌辨識成為其中一個重要的議題，業已吸引了許多電腦視覺相關研究團隊的投入，目前也發展出多種不同的車牌定位方法。然而在車牌識別的作業過程中，最關鍵也是最困難的步驟是車牌定位。因為在車牌的偵測程序裡，不僅會受到外在環境的影響，各個國家地區車牌設計上的差異也增加其困難度。
目前多數的研究方法會有一些前提假設，在面臨較惡劣的外在環境，通常辨識效能會無法因應狀況而受到影響。因此，本研究提出一個有效的前置處理演算法，能降低外在環境對車牌的影像品質所造成之影響，同時提供較佳的車牌候選區域與候選車牌的產生，再透過多重解析度來分析影像，並從中提取碎形維度特徵，並藉由這些特徵來訓練分類器。目的是讓系統在遭遇各種環境狀況下，能藉由學習的方式來適應新的環境。
本研究實驗結果顯示，對於不同的擷取角度、不同的車牌樣式、不同的光源條件之汽車影像，能完美的定位出車牌位置，而疑似車牌之非車牌樣本，亦能有效的成功拒絕。

Due to the rapid growth of developing countries, the demand of vehicles increased astonishingly. In dealing with the complexity of transportation management, license plate recognition has become an important topic, in which many researches were involved particularly in the field of computer vision. In the process of license plate recognition, the most critical and difficult operation is locating the positioning of a license plate. Not only interfered by external environmental factors, the difference in license plate design of the various countries and regions accumulates the challenges.
Many proposed methods have preset restrictions. When encountering relatively poor external environment, some recognition systems could not cope condition effectively. In this study, an efficient pre-processing algorithms is introduced in order to reduce the external interference on the image quality of license plates. Meanwhile candidate-positions of license plate is generated. Thereafter, a multi-resolution analyze is activated and fractal feature vector will be extracted to train classifier. That simply allows the system to adapt to all changes.
The present study shows, under different angles, different formats and different lighting conditions, the proposed method is able to identify license plate location errorless, and reject non-license plate image nearly perfect.

摘要 i
Abstract ii
誌謝 iii
目錄 iv
圖目錄 vi
表目錄 viii
附錄 ix
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機 3
1.3 研究目的 3
第二章 文獻探討 5
2.1 邊界與線條特徵 5
2.2 全域影像特徵 7
2.3 紋理特徵 9
2.4 色彩特徵 9
2.5 字元特徵 11
第三章 研究方法 12
3.1 影像前置處理 13
3.2 車牌候選區域和候選車牌的產生 14
3.3 影像特徵擷取 (Feature Extraction) 15
3.3.1 多重解析度分析(Multi-resolution Analysis) 15
3.3.2 碎形維度的特徵值計算(Fractal Dimension Vectors) 16
3.3.3 特徵選取 16
3.4 分類器 16
3.4.1 常見的分類器 17
3.4.1.1 MDC 17
3.4.1.2 Bayes classifier 17
3.4.1.3 Fuzzy KNN 18
3.4.1.4 PNN 18
3.4.1.5 SVM 19
3.5 車牌定位 20
3.5.1 合併候選車牌 20
3.5.2 候選車牌切割 22
3.5.3 候選車牌驗證 23
第四章 實驗結果 25
4.1 實驗環境 25
4.2 實驗影像 25
4.3 實驗結果 26
4.3.1 前置處理 26
4.3.2 車牌候選區域與候選車牌 29
4.3.3 分類器的訓練與測試 31
4.3.4 候選車牌的合併、切割與驗證 33
第五章 結論 39
參考文獻 40

[1]Shan Du, Mahmoud Shehata and Wael Badwy, “Automatic License Plate Recognition(ALPR)：A State-of-the-Art Review,” IEEE Trans. Circuit and Systems for Video Technology., vol. 23, no. 2, pp.311-325, Feb. 2013
[2]台灣歷年車輛牌照
http://zh.wikipedia.org/wiki/%E8%87%BA%E7%81%A3%E8%BB%8A%E8%BC%9B%E7%89%8C%E7%85%A7
[3]現行車牌，新式與舊式車牌規格之說明http://www.thb.gov.tw/TM/Files%5CWebpage%5C201211%5C13_101HC08488-1.pdf
[4]遠通電收之爭議
http://zh.wikipedia.org/wiki/%E9%81%A0%E9%80%9A%E9%9B%BB%E6%94%B6
[5]ETC時事議題-辨識系統探討
https://tw.autos.yahoo.com/news/%E7%B0%A1%E5%96%AE%E4%B8%8A%E8%B7%AF-%E9%AB%98%E5%85%AC%E5%B1%80-%E5%8F%AF%E4%BB%A5%E4%B8%8D%E8%A6%81%E5%97%8E-023147625.html
[6]遠通電收對其辨識系統說明
http://www.fetc.net.tw/service/qa/topic.html
[7]高速公路局對ETC之辨識系統說明
http://www.freeway.gov.tw/Publish.aspx?cnid=195&p=4882
[8]V. Kamat and S. Ganesan, “An Efficient Implementation of the Hough Transform for Detecting Vehicle License Plates Using DSP''S,” Real-Time Technology and Applications Symposium., pp.58-59, May 1995
[9]Tran Duc Duan, Duong Anh Duc and Tran Le Hong Du , “Combining Hough Transform and Contour Algorithm for Detecting Vehicles License-Plates,” Int. conf. Intelligent Multimedia, Video and Speech Processing, pp.747-750, Oct. 2004
[10]M J. Ahmed, M Sarfaz, A. Zidouri and K G. AI-Khatib, “License Plate Recognition System,” IEEE Int. Conf. Electronics, Circuits and System,pp.898-901, Dec. 2003
[11]M. Sarfraz, M. J. Ahmed and S. A. Ghazi, “Saudi Arabian License Plate Recognition System,” Int. Conf. Geom. Model. Graph., pp. 36–41, 2003
[12]Danian Zheng, Yannan Zhao and Jiaxin Wang, “An efficient method of license plate location,” Pattern Recognition Letters, vol. 26, pp.2431-2438, 2005
[13]S. Wang and H. Lee, “Detection and Recognition of License Plate Characters with Different Appearances,” Int. Conf. Intell. Transp. Syst., vol. 2, pp.979–984, 2003
[14]B.-F. Wu, S.-P. Lin and C.-C. Chiu, “extracting characters from real vehicle license plates out of doors,” IET Comput. Vis., vol. 1, no. 1,pp. 2–10, 2007
[15]Ying Wen, Yue Lu, Jingqi Yan, Zhenyu Zhou, Karen M. von Deneen and Pengfei Shi, “An Algorithm for License Plate Recognition Applied to Intelligent Transportation System,” IEEE Trans. Intelligent Transportation System., vol. 12, no. 3, pp.830-844 Sept. 2011
[16]K. Deb, H.-U. Chae and K.-H. Jo, “Vehicle License Plate Detection Method Based on Sliding Concentric Windows and Histogram,” J. Comput., vol. 4, no. 8, pp. 771–777, 2009.
[17]RGB色彩模型
http://en.wikipedia.org/wiki/RGB_color_model
[18]HSL與HSV色彩模型
http://en.wikipedia.org/wiki/HSL_and_HSV
[19]X. Shi, W. Zhao and Y. Shen, “Automatic License Plate Recognition System Based on Color Image Processing,” Lecture Notes Comput. Sci., vol.3483, pp.1159-1168, 2005
[20]E. R. Lee, P. K. Kim and H. J. Kim, “Automatic Recognition of a Car License Plate Using Color Image Processing,” IEEE Int. Conf. Image Process., vol. 2. pp.301-305, Nov. 1994
[21]S. Yohimori, Y. Mitsukura, M. Fukumi, N. Liu and N. Pedrycz, “License plate detection system by using threshold function and improved template matching method,” IEEE Annu. Meeting Fuzzy Inform., vol. 1, pp.357-362, Jun. 2004
[22]S. Draghici, “A Neural Network Based Artificial Vision System for License Plate Recognition,” Int. J. Neural System., vol. 8, no. 1 , pp.113-126, Dec. 1997
[23]H. Hontani and T. Koga, “Vehicle Plate Recognition for Wireless Traffic Control and Law Enforcement System,” IEEE Int. Conf. Ind. Tech., pp.1800-1804, 2006
[24]Wen-Li Lee, Yung-Chang Chen and Kai-Sheng Hsieh, “Ultrasonic Liver Tissues Classification by Fractal Feature Vector Based on M-Band Wavelet Transform,” IEEE Trans. Medical Imaging., vol. 22, pp.382-392, 2003
[25]Alex P. Pentland, “Fractal-Based Description of Natural Scenes,” IEEE Trans. Pattern Analysis and Machine Intelligence., vol. 6, pp.661-667, 1984
[26]Wen-Li Lee and Kai-Sheng Hsieh, “An off-line segmentation system for texture analysis of liver ultrasonic images, ”J. Information Technology in Healthcare, vol. 4, pp.301-307, 2006
[27]Wen-Li Lee and Kai-Sheng Hsieh, “A robust algorithm for the fractal dimension of images and its applications to the classification of natural images and ultrasonic liver images,” Signal Processing., vol. 90, pp.1894-1904, Jun. 2010
[28] Sergios Theodoridis and Konstantinos Koutroumbas, Pattern Recognition, forth edition, 2008
[29]M. Nadler and E.P.Smith, Pattern Recognition Engineering. New York：Wiley，1993
[30]J. M. Keller, M.R.Gray and J. A. Givens, “A Fuzzy K-Nearest Neighbor Algorithm” IEEE Trans. Systems, Man and Cybernetics., vol. 23, smc. 15, no. 4, pp.580-585, Jul. 1985
[31]D. F. Specht, “Probabilistic Neural Network,” Neural Networks., vol. 3, no. 1, pp.109-118, Jan. 1990
[32]A Library for Support Vector Machines
http://www.csie.ntu.edu.tw/~cjlin/libsvm/