臺灣博碩士論文加值系統

本論文探討汽車車牌辨識研究，並針對台灣在 2012 年所發行的新式汽車車牌提出一個新的字元辨識方法。而提出的方法中包含車牌定位、車牌字元切割及車牌字元切割，車牌定位的部份是應用開放碼 OpenALPR，定位出車牌矩形的四點並計算 Homography matrix 來正規化車牌影像。字元切割方法是基於字元模組來設計結合字元投影輪廓及已知字元寛度的混合法，可以有效切割出字元影像。利用散射變換 (Scattering Transform) 演算法來擷取車牌字母之特徵值，並使用支持向量機 (SVM, Support Vector Machine) 模組做分類來辨識字元。為訓練 SVM，我們發展能模擬車牌取像及車牌定位誤差的影像合成系統，利用合成的擬真字母影像做為訓練集。完成的車牌辨識系統對於車牌辨識有不錯的效果，並能抵抗字元影像一定程度上的型態變化，平均辨識率達 98.82%。

In this thesis, we proposed a new method for recognizing the Taiwanese car license plates issued since 2012.The proposed method consists of a license plate detection step, a license plate rectification step, a character segmentation step, a feature extraction step, and a character recognition step. The license plate detection step is accomplished by using the OpenALPR, an open source license plate recognition software package.The four corners of the detected license plate are used to compute a homography matrix to rectify the license plate image. A model-based character segmentation method is developed to obtain single-character
images. Features of the character images are extracted with scattering transform. Characters are recognized using the support vector machine (SVM). In order to train the SVM, a license plate image synthesis system is implemented to provide the training images. The image synthesis system generates license plate images corrupted by several effects such as out-of-focus blur, shadow, image noise, and character deformation due to the license plate localization error. Real experiments show that the average accuracy of the proposed method is 98.82%. It even outperforms the convolutional neural network approach in digits recognition.

目 次
致謝 ................................................................................................................................... I
摘要.................................................................................................................................... II
Abstract ............................................................................................................................. III
目 次 .................................................................................................................................. IV
圖 目 次 ............................................................................................................................. VI
表 目 次 ............................................................................................................................. VIII
第一章 緒論...................................................................................................................... 1
1.1 研究動機 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 相關文獻探討 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 車牌定位技術 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 車牌字元切割 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.3 車牌字元辨識 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 研究目標 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 論文架構 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
第二章 研究方法.............................................................................................................. 5
2.1 車牌種類介紹 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 計算流程 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 辨識用字元影像合成 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 特徵抽取 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.1 CNN 概述 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4.2 散射變換 (Scattering Transform) 概述 . . . . . . . . . . . . . . . 19
第三章 實驗過程與結果.................................................................................................. 22
3.1 實驗過程 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.1 實驗平台 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.2 實驗一：車牌定位的角度範圍 . . . . . . . . . . . . . . . . . . . 23
3.1.3 實驗二：測試資料集 . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.4 SVM 及字元辨識 . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2 實驗結果 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
第四章 結論與未來方向.................................................................................................. 34
4.1 結論 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.2 未來方向 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
參 考 文 獻 ........................................................................................................................ 36
圖 目 次
1.1 各國車牌圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 澳洲字體 (a) 英文模組, (b) 數字模組 . . . . . . . . . . . . . . . . . . . . 6
2.2 系統計算流程圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Homography 示意圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 (a) 車牌原始影像, (b) 車牌偵測影像, (c) Homography 的車牌影像 . . . . 10
2.5 (a) 字元投影, (b) 車牌字元切割示意圖 1 . . . . . . . . . . . . . . . . . . 10
2.6 (a) 字元投影, (b) 車牌字元切割示意圖 2 . . . . . . . . . . . . . . . . . . 11
2.7 (a) 字元投影, (b) 字元切割失敗圖 1 . . . . . . . . . . . . . . . . . . . . 11
2.8 (a) 字元投影, (b) 字元切割失敗圖 2 . . . . . . . . . . . . . . . . . . . . 12
2.9 合成字元流程圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.10 陰影化示意圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.11 陰影化範例 (a) 英文 C, (b) 數字 6 . . . . . . . . . . . . . . . . . . . . . . 15
2.12 高斯雜訊範例 (a) 英文 L, (b) 數字 6 . . . . . . . . . . . . . . . . . . . . 16
2.13 模糊化範例 (a) 英文 L, (b) 數字 6 . . . . . . . . . . . . . . . . . . . . . . 16
2.14 (a) 英文字元範例, (b) 數字字元範例 . . . . . . . . . . . . . . . . . . . . 17
2.15 CNN 結構圖 (copyright from LeNet) . . . . . . . . . . . . . . . . . . . . 19
2.16 (a) 字型變化, (b) 局部平移描述變形示意 . . . . . . . . . . . . . . . . . 19
2.17 ST 結構示意圖 [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 取像設備示意圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2 車牌偵測角度示意圖 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 (a) 拍攝角度過大, (b) 拍攝距離太遠 . . . . . . . . . . . . . . . . . . . . 25
3.4 (a) 偵測後所得車牌影像, (b) Homography 後的車牌影像 . . . . . . . . . 25
3.5 (a) 偵測後所得車牌影像, (b)Homography 後的失敗影像 . . . . . . . . . 26
3.6 (a) 車牌影像, (b) 字元切割錯誤的範例 1 . . . . . . . . . . . . . . . . . . 32
3.7 (a) 車牌影像, (b) 字元切割錯誤的範例 2 . . . . . . . . . . . . . . . . . . 32
3.8 (a) 車牌影像, (b) 字元切割錯誤的範例 3 . . . . . . . . . . . . . . . . . . 32
3.9 (a) 車牌影像, (b) 陰影造成字元誤判 . . . . . . . . . . . . . . . . . . . . 33
3.10 (a) 車牌影像, (b) 字元影像被切割到邊緣 . . . . . . . . . . . . . . . . . 33
3.11 (a) 車牌影像, (b) 字元誤判範例 1 . . . . . . . . . . . . . . . . . . . . . . 33
3.12 (a) 車牌影像, (b) 字元誤判範例 2 . . . . . . . . . . . . . . . . . . . . . . 33
表 目 次
2.1 台灣車牌樣式 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1 車牌定位角度測試 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 SVM 模組辨識率 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.3 L2RLR-P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.4 L2RL2L-D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.5 L2RL2L-P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.6 L2RL1L-D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.7 C & S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.8 L1RL2L . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.9 L1RLR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.10 L2RLR-D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.11 辨識率比較 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.12 扣除字元切割錯誤後的辨識率 . . . . . . . . . . . . . . . . . . . . . . . 33

[1] J.BrunaandS.Mallat, “Invariantscatteringconvolutionnetworks,” IEEETransactions on, Pattern Analysis and Machine Intelligence, vol. 35, pp. 1872–1886, Aug 2013.
[2] S. Du, M. Ibrahim, M. Shehata, and W. Badawy, “Automatic license plate recognition (alpr): A state-of-the-art review,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 23, pp. 311–325, Feb 2013.
[3] 吳瑞展 and 吳先晃, “多影幅與及時低解析度道路視訊車牌辨識,” 2007.
[4] H.BaiandC.Liu,“A hybrid license plate extraction method based on edge statistics and morphology,” in International Conference on Pattern Recognition, vol. 2, pp. 831–834, Aug 2004.
[5] L.Luo, H.Sun, W.Zhou, and L.Luo, “Ane fficient method of license plate location,” in International Conference on Information Science and Engineering, pp. 770–773, Dec 2009.
[6] S.-Z. Wang and H.-J. Lee, “Detection and recognition of license plate characters with different appearances,” in Intelligent Transportation Systems, vol.2, pp.979–984 vol.2, Oct 2003.
[7] F. Faradji, A.H.Rezaie, and M. Ziaratban, “A morphological-based license plate location,” in Image Processing,IEEE International Conference, vol. 1, pp. I – 57–I – 60, Sept 2007.
[8] K. Kanayama, Y. Fujikawa, K. Fujimoto, and M. Horino, “Development of vehicle license number recognition system using real-time image processing and its application 36 to travel-time measurement,” in IEEE Vehicular Technology Conference, pp. 798–804, May 1991.
[9] N.Bellas, S.Chai, M.Dwyer and D.Linzmeier,“Fpga implementation of a license plate recognition soc using automatically generated streaming accelerators,” in Parallel and Distributed Processing Symposium, pp. 8 pp.–, April 2006.
[10] H.-H. P. Wu, H.-H. Chen, R.-J. Wu, and D.-F. Shen, “License plate extraction in low resolution video,” in Pattern Recognition, pp. 824–827, 2006.
[11] M. I. C. M. and A. Z. S., “License plate location based on a dynamic pcnn scheme,” in Neural Networks,International Joint Conference, pp. 1195–1200 vol.2, July 2003.
[12] K.Miyamoto, K.Nagano, M.Tamagawa, I.Fujita and M.Yamamoto,“Vehicle license plate recognition by image analysis,” in Industrial Electronics, Control and Instrumentation, pp. 1734–1738 vol.3, Oct 1991.
[13] H. ke Xu, F. hua Yu, J. hua Jiao, and H. sheng Song, “A new approach of the vehicle license plate location,” in Parallel and Distributed Computing, Applications and Technologies, pp. 1055–1057, Dec 2005.
[14] R. Zunino and S. Rovetta, “Vector quantization for license-plate location and image coding,”IEEETransactionsonVehicularTechnology,vol.47,no.1,pp.159–167,2000.
[15] C. Anagnostopoulos, I. Anagnostopoulos, V. Loumos, and E. Kayafas, “A license plate-recognition algorithm for intelligent transportation system applications,” Intelligent Transportation Systems, IEEE Transactions on, vol. 7, pp. 377–392, Sept 2006.
[16] X. Shi, W. Zhao, and Y. Shen, “Automatic license plate recognition system based on color image processing,” in Computational Science and Its Applications,
pp. 1159–1168, 2005.
[17] S.-L. Chang, L.-S. Chen, Y.-C. Chung, and S.-W. Chen, “Automatic license plate recognition,” IEEE Transactions on Intelligent Transportation Systems, vol. 5, no. 1, pp. 42–53, 2004.
[18] J. Matas and K. Zimmermann, “Unconstrained licence plate and text localization and recognition,” in Intelligent Transportation Systems, pp. 225–230, Sept 2005.
[19] Z.-X. Chen, C.-Y. Liu, F.-L. Chang, and G.-Y. Wang, “Automatic license-plate location and recognition based on feature salience.,” IEEE T. Vehicular Technology, vol. 58, no. 7, pp. 3781–3785, 2009.
[20] Y.Wang, H.Zhang, X.Fang, and J.Guo, “Low-resolution chinese character recognition of vehicle license plate based on albp and gabor filters,” in International Conference on Advances in Pattern Recognition, pp. 302–305, Feb 2009.
[21] K. Kanayama, Y. Fujikawa, K. Fujimoto, and M. Horino, “Development of vehicle license number recognition system using real-time image processing and its application to travel-time measurement,” in 41st IEEE Gateway to the Future Technology in Motion on Vehicular Technology Conference, pp. 798–804, May 1991.
[22] T. D. Duan, T. L. H. Du, T. V. Phuoc, and N. V. Hoang, “Building an automatic vehicle license-plate recognition system,” in RIVF, pp. 59–63, February 2005.
[23] Q. Gao, X. Wang, and G. Xie, “License plate recognition based on prior knowledge,” in IEEE International Conference on Automation and Logistics,, pp. 2964–2968, August 2007.
[24] J.-M. Guo, I. Member, and Y.-F. Liu, “License plate localization and character segmentation with feedback self-learning and hybrid binarization techniques,”in IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, vol. 57, pp. 1417–1424, May 2008.
[25] K.-B. Kim, S.-W. Jang, and C.-K. Kim, “Recognition of car license plate by using dynamical thresholding method and enhanced neural networks,” in Computer Analysis of Images and Patterns, pp. 309–319, August 2003.
[26] A.Capar and M.Gokmen,“Concurrent segmentation and recognition with shape-driven fast marching methods,” in IEEE International Conference on Pattern Recognition, vol. 1, pp. 155–158, 2006.
[27] S. Nomura, K. Yamanaka, O. Katai, H. Kawakami, and T. Shiose, “A novel adaptive morphological approach for degraded character image segmentation,” in IEEE International Conference on Pattern Recognition, vol. 38, 2005.
[28] S. Nomura, K. Yamanaka, O. Katai, and H. Kawakami, “A new method for degraded color image binarization based on adaptive lightning on grayscale versions,” in Institute of Electronics, Information and Communication Engineers, vol. 87-D, pp. 1961–1975, 2004.
[29] S.Ozbay and E.Ercelebi,“Automatic vehicle identification by plate recognition,”2005.
[30] P. Hu, Y. Zhao, Z. Yang, and J. Wang, “International conference on recognition of gray character using gabor filters,” in Information Society Innovation Fund, vol. 1, pp. 419–424, July 2002.
[31] S. N. H. S. Abdullah, M. Khalid, and R. Yusof, “License plate recognition using multicluster and multilayer neural networks,” in Information and Communication Technologies, vol. 1, pp. 1818–1823, 2006.
[32] C. J. Burges, “A tutorial on support vector machines for pattern recognition,” Data Mining and Knowledge Discovery, vol. 2, pp. 121–167, 1998.
[33] F. K, “Neocognitron: A self-organizing neural network for a mechanism of pattern recognition unaffected by shift in position,” in Biological Cybernetics, pp. 193–202, 1980.
[34] D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” in International Journal of Computer Vision, pp. 552–555, January 2004.
[35] A. Graps, “An introduction to wavelets,” IEEE Computational Science Engineering, vol. 2, pp. 50–61, Summer 1995.
[36] R.-E. Fan, K.-W. Chang, C.-J. Hsieh, X.-R. Wang, and C.-J. Lin, “LIBLINEAR: A library for large linear classification,” Journal of Machine Learning Research, vol. 9, pp. 1871–1874, 2008.