臺灣博碩士論文加值系統

一個車牌辨識系統是用來辨識交通工具上之牌照號碼及英文字母，以供安全系統、
停車場管理，甚至智慧型運輸系統之用。本論文所討論之系統是使用影像處理技術來
進行車牌偵測、定位、分割、以及英文字母和阿拉伯數字辨識，之後將所得資訊儲存
以供後續之用。
本系統首先拍攝汽車之數位影像，之後在數位影像中使用邊緣偵測技術把交通工
具之牌照區域擷取下來。接著再使用聯通元件分析來分割牌照上之字元，最後利用影
像慣量不變方法來辨識阿拉伯數字和英文字母，處理所得資訊可以用來和資料庫所存
資料相比對。本系統在MATLAB® 平台上實踐及模擬，並且以實際影像測試其效能，
實驗結果證明本系統可在相當高的準確度下，快速、有效率地辨識汽車車牌。實驗結
果證明本論文所提出方法可以快速、有效率地辨識汽車車牌，達到95.55%成功率的
準確度。

A license plate recognition (LPR) system is designed to recognize the alphabets and numbers on the plates of vehicles, and its goals are for security, parking lot management, and intelligent transportation, etc. The proposed LPR system is based on image processing techniques to detect, locate, and segment the license plate in a digital input image, and recognize the alphabets and numbers on the plate. The acquired information by the LPR system is stored for further applications. The proposed method first captures the digital image of an automobile. Vehicle number plate region is extracted using the edge detection in an image. Connected component analysis (CCA) method is used to detect the plate and segmented characters on it. The moment invariants technique is used for the character recognition. The resulting data is then used to compare with the records on a database. The system is implemented and simulated in MATLAB® , and its performance is tested on real image. Experimental results proved that the proposed method can be used for fast and efficient recognition of the car license plates with 95.55% successes rate accuracy.

摘要 …………………………………….…………………………………………………..i
Abstract .............................................................................................ii
Acknowledgement ………………………………………………………………………..iii
Contents …………………………………………………………………………………...iv
Tables content ............................................................................................ vii
Figures content ..........................................................................................vi i i
Chapter 1 Introduction ………………………………………...……………………….1
1.1. Applications ………………………………………………………………………3
Chapter 2 Image Processing Technique and Benchmark LPR Methods …..………..7
2.1 Optical Character Recognition by Template Matching …………………………...8
2.2 Artificial Neural Networks Technique …………………………………………..10
2.3 Morphological Operations ……………………………………………….………13
2.4 Connected Component and Clustering Techniques ……………………………..14
2.5 Edge Detection …………………………………………………………………..14
2.5.1 The Classical Edge Detection Operator …….…………....………………15
2.5.2 Roberts Edge Operator ……………………………...……………………16
2.5.3 Sobel Operator …………...…………………………………….…………16
2.5.4 Prewitt Operator ………………………………………………………….16
2.5.5 Krisch Operator …………………………………………………………..17
2.5.6 Canny Edge Detection Operator ………………………………………….17
2.6 Literature Review and Comments ……………………………………………….17
Chapter 3 The Proposed Method ……………………………………………………..20
3.1 Image Capture and Pre-processing ……………………………...………….……20
3.1.1 Otsu’s Method ……………….…………………………………...........…22
3.1.2 Thresholding ……………….…………………………………………..…23
3.1.3 Binarization ………………....……………………………………………24
3.2 Plate Region Extraction …………………………………………….……………25
3.2.1 License Plate Detection ….......................................................................29
a) Edge Detection ………………………………………………………………29
a.1 Gaussian Filter …………..………………………………………………….30
a.2 Finding the Intensity Gradient of the Image …...………………….………..30
a.3 Non-Maximum Suppression ………..………………………………………31
a.4 Double Treshold ………………..…………………………………………..32
a.5 Edge Tacking by Hsteresis ………...………………………………………..32
b) Plate Extraction ……………………………………………………………..33
b.1 Vertical Edge Extraction ……..……………………………….……………33
b.2 Vertical Edge Detection Algorithm ……………..………………………….34
c) Size Normalization ………….………………………………….……………35
3.3 Segmentation …………………………………………………….………………36
3.3.1 Character Segmentation ………………………………………………….38
a) Tilt Correction of the Plate ………………………………..…………………38
b) Median Filtering For Noise Reduction ……………………………………...39
c) The Process of Segmentation …………………………………..……………40
3.4 Recognition …………………………………………………….……………..…45
3.4.1 Definition …………………………………………………………………45
3.4.2 Categories of Invariant …………...………………………………………47
3.4.3 Moments ………………………………………………………………….48
a) Geometric and Complex Moments ………………………………………….49
b) Orthogonal Moments ………………………………………………………..50
3.4.4 Moment Invariants to Translation, Rotation and Scaling ……………..….……52
a) Invariants to Translation ……………………………………………………..52
b) Invariants to Uniform Scaling ……………………………………………….53
c) Traditional Invariants to Rotation ………………………………...…………54
3.4.5 Rotation Invariants from Complex Moments ………………………………….55
a) Construction of Rotation Invariants …………………….……………..…….55
3.4.6 Rotation Invariants for Recognition of Symmetric Objects ……………….…..56
3.4.7 Use Orthogonal Moments ……………………………………………………..56
Chapter 4 Experimental Result …………………………….…………………………60
4.1 Database …………………………...…………………………………….………62
4.2 Software ……………………………………………………………………….…64
4.3 Difficulties ……………………………………………………………………….66
Chapter 5 Conclusion ………………………………………………………………….68
Appendix A …………………………………………………………………….…………70
Appendix B …………………………………………………………………………….…76
References ………………………………………………………………………………..86
Tables content
Table 1.1 Compare existing methods …………………………………………………….19
Table 4.1 Training data: C (characters), A (amount) ……………………………………..63
Table 4.2 Response time for algorithm ………………..……………………….…………65
Table 4.3 Evaluation of algorithm result …………………………………………………65
Table 4.4 Analysis of each part ……………….………………………………………….65
Figures content
Fig. 1.1 LPR system ………………………………………………………………………..6
Fig. 2.1 Three main parts of LPR…………………………………………………………...9
Fig. 2.2 Block diagram of the ANNs recognition system ………………………………...12
Fig. 3.1. Flow diagram of number plate recognition algorithm ………………………......20
Fig. 3.2 (a) Original image, (b) Gray-level ……………………………………………….21
Fig. 3.3 Histogram chart of gray-level ……………………………………………………21
Fig. 3.4 (a) Original image, (b) Image thresholded using Otsu's algorithm ……………...22
Fig. 3.5 (a) Original image, (b) Image after threshold…………………………………….23
Fig. 3.6 (a) Original image, (b) Binarization image ……………………….……………...25
Fig. 3.7 Plate location procedure ………………………………………………………….27
Fig. 3.8 Black value in image tested ……………………………………………………...34
Fig. 3.9 Converting center pixel with 4 angles …………………………………………...34
Fig. 3.10 Find and extract target plate …………………………………………………….35
Fig. 3.11 Plate segmented ………………………………………………………………...38
Fig. 3.12 The result of median filtering …………………………………………………..40
Fig. 3.13 Plate segmentation procedure ……………………………………..................... 41
Fig. 3.14(a) Skewed plate (b) Put bounding box (c) Calculate skew angle θ (d) Result after
run this program …………………………………………………………………………..43
Fig. 3.15 Characters of segmentation……………………………………………………...44
Fig. 4.1 Two kinds of Taiwanese car plates ………………………………………………60
Fig. 4.2 Alphabetical samples …………………………………………………………….62
Fig. 4.3 Numbers sample ………………………………………....................................... 63
Fig. 4.4 Samples of car’s images ………………………………........................................66
Fig. A-1 Convert to graylevel and binarization ………………………............................. 70
Fig. A-2 Find target and segmentation ……………………………………………………70
Fig. A-3 Recognition and moment function ……...……………………………………….71
Fig. A-4 Calculated moment and show result …………………………………………… 71
Fig. A-5 Binarization function …...……………………………………………………….72
Fig. A-6 Edge detection function …………………………………………………………72
Fig. A-7 Feature function ………...……………………………………………………….73
Fig. A-8 Recognition function …………………………………………………………………………..73
Fig. A-9 Invariable moment function …………………………………………………......74
Fig. A-10 All run and result ……………………………………………………………….75
Fig. B-1 Plate localization error ………………………………………………………..…76
Fig. B-2 Characters segmentation errors………………………………………………..…79
Fig. B-3 Characters recognition errors …………………………………………………....85

[1] D. Renuka Devi, and D. Kanagapushpavalia, “Automatic License Plate Recognition,” 3rd
International Conference on Trendz in Information Scinces & Computing, Pages 75-
78, (2011)
[2] Kuo-Ming Hung, and Ching-Tang Hsieh “A Real-Time Mobile Vehicle License Plate
Detection and Recognition,” Tamkang Journal of Science and Engineering, Vol. 13, No. 4,
Pages 433- 442, (2010)
[3] S. Chang, L. Chen, Y. Chung, and S. Chen, “Automatic License Plate Recognition,”
IEEE Transactions on Intelligent Transportation Systems, Vol. 5, Pages 42- 53, (2004)
[4] Hadi Sharifi Kolour, “An Evaluation of License Plate Recognition Algorithms,”
International Journal of Digital Information and Wireless Communications (IJDIWC), Pages
247- 253, (2011)
[5] Er. Kavneet Kaur, and Vijay Kumar Banga, “Number Plate Recognition Using OCR
Technique,” International Journal of Research in Engineering and Technology (IJRET), Vol.
2, Pages 286- 290, (2013)
[6] Anu Agarwal, and Sudhir Goswami, “An Efficient Algorithm for Automatic Car Plate
Detection & Recognition,” 2nd International Conference on Computational Intelligence &
Communication Technology, Pages 644- 648, (2016)
[7] H. Erdinc Kocera, and K. Kursat Cevik, “Artificial Neural Networks based Vehicle
License Plate Recognition,” Procedia Computer Science, Pages 1033- 1037, (2011)
[8] Xiaojun Zhai, Faycal Benssali, and Soodamani Ramalingam, “License Plate Localization
based on Morphological Operations,” 11th International Conference on Control, Automation,
Robotics, and Vision, Pages 7- 10, (2010)
[9] Abdollah Amirkhani Shahraki, Amir Ebrahimi Ghahnavieh, and Seyyed Abdollah
Mirmahdavi, “License Plate Extraction from Still Images,” 4th International Conference on
Intelligent Systems, Modelling, and Simulation, Pages 45- 48, (2013)
[10] V. Karthikeyan, V. J. Vijayalakshmi, and P. Jeyakumar, “License Plate Segmentation
Using Connected Component Analysis,” Journal of Electronics and Communication
Engineering (IOSR-JECE), Vol. 4, Pages 18- 24, (2013)
[11] Lejiang Guo, Yahui Hu, and Xuanlai Tang, “The Edge Detection Operators and Their
Application in License Plate Recognition,” IEEE, (2010)
[12] C. Chunyu, C. Baozhi, C. Xin, and Z. Chen, “Application of Image Processing to the
Vehicle License Plate Recognition,” Proceedings of the 2nd International Conference on
Computer Science and Electronics Engineering (ICCSEE), Pages 2867- 2869, (2013)
[13] Wei Xie, and Y. Wu, “License Plate Automatic Recognition System Based on
MATLAB-GUI,” The Open Automation and Control Systems Journal, Pages 497- 502,(2014)
[14] Atul Kumar, and Sunila Godara, “A Review: On Number Plate Recognition,”
International Journal of Science and Research (IJSR), Vol. 4, Pages 1964- 1967, (2015)
[15] D. Liu, and J. Yu, “Otsu method and k-means,” 9th International Conference on Hybrid
Intelligent Systems, Vol. 1, Pages 344- 349, (2009)
[16] M. Sezgin, and B. Sankur, "Survey Over Image Thresholding Techniques and
Quantitative Performance Evaluation," Journal of Electronic Imaging, Pages 146- 165, (2013)
[17] Yu Chen, Dian-ren Chen, and Yang Lei Li, “Otsu's Thresholding Method Based on
Gray Level-Gradient Two-Dimensional Histogram,” 2nd International Asia Conference on
Information in Control, Automation, and Robotics, Vol. 3, Pages 282- 285, (2010)
[18] Rongbao Chen, and Yunfei Luo, “An Improved License Plate Location Method Based
on Edge Detection,” ELSEVIER, Physics Procedia 24, Pages 1350- 1356, (2012)
[19] Xiao Juan Ma, Renlong Pan, and Lin Wang, “License Plate Character Recognition
Based on Gaussian Hermite Moments,” Education Technology and Computer Science, Vol.
3, Pages 11- 14, (2010)
[20] Lei Wang, and Hanli Wang, “License Plate Recognition Method Based on Double-Edge
Detection,” Computer Engineering and Applications, Pages 169- 173, (2012)
[21] S. C. Pei, and C. N. Lin, “Image Normalization for Pattern Recognition,” Image Vis
Computing, Vol. 13, No. 10, Pages 711- 723, (2002)
[22] J. M. Guo, and Y. F. Liu, “License Plate Localization and Character Segmentation with
Feedback Self-Learning and Hybrid Binarization Techniques,” IEEE Transaction on
Vehicular Technology, Vol. 57, No. 3, Pages 1417- 1424, (2012)
[23] Lihong Zheng, and Xiangjian He, “Efficient Character Segmentation on Car License
Plates,” 11th International Conference on Control Automation Robotics, and Vision
(ICARCV), Pages 1139- 1144, (2010)
[24] F. Kahraman, B. Kurt, and M. Gokmen. “License Plate Character Segmentation Based
on the Gabor Transform and Vector Quantization,” International Symposium on Circuits and
Systems, (ISCAS), Pages 381- 388, (2003)
[25] A. Conci, J. E. R. de Carvalho, and T. W. Rauber, “A Complete System for Vehicle
Plate Localization, Segmentation and Recognition in Real Life Scene,” IEEE Latin America
Transactions, Vol. 7, No. 5, Pages 497- 506, (2009)
[26] Yue Li, Asok Ray, and Thomas A. Wettergren, “Target Detection and Target Type &
Motion Classification and Comparison of Feature Extraction Algorithms,” American
Control Conference, Pages 1132- 1137, (2014)
[27] X. Dai, H. Zhang, H. Shu, and L. Luo, “Image Recognition by Combined Invariants of
Legendre Moment,” IEEE, International Conference, Pages 1793- 1798, (2010)
[28] J. Flusser, T. Suk, and B. Zitova, Moments and Moment Invariants in Pattern
Recognition, Wiley & Sons Publications Ltd, (2009)
[29] E. X. Huang, and A. K. Fung, “An Application of Sampling Theorem to Moment
Method Simulation in Surface Scattering,” Journal of Electromagnet, Waves, and Appl, Vol.
20, No. 4, Pages 531- 546, (2006)
[30] Khalid Mohamed Hosny, “Accurate Orthogonal Circular Moment Invariants of Gray-
Level Images,” Journal of Computer Science, Pages 715- 722, (2011)
[31] A. Stern, I. Kruchakov, E. Yoavi, and S. Kopeika, “Recognition of Motion-Blurred
Images by Use of the Method of Moments,” Applied Optics, Vol. 41, Pages 2164- 2172,
(2006)
[32] Meng Yin Fu, Fang Yu Liu, and Mei ling Wang, “Background Pixels Mutation
Detection and Hu Invariant Moments Based Traffic Signs Detection on Autonomous
Vehicles,” Chinese Control Conference, Pages 670- 674, (2014)
[33] Ming-Kuei Hu, “Visual Pattern Recognition by Moment Invariants,” IRE Transactions
on Information Theory, Vol. 8, No. 2, Pages 179- 187, (1962)
[34] Huibao Lin, Jennie Si, and Glen P. Abousleman, “Orthogonal Rotation-Invariant
Moments for Digital Image Processing,” IEEE Transactions on Image Processing, Vol. 17,
Pages 272- 282, (2008)
[35] Yaser. S. Mostafa, and Demetri Psaltis, “Image Normalization by Complex Moments,”
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 7, No. 1, Pages 46- 55, (2005)
[36] K. M. Hosny, “On the Computational Aspects of Affine Moment Invariants for Gray-
Scale Images,” Vol. 195, Pages 762- 771, (2008)
[37] Qian Li, Hongqing Zhu, and Qian Liu, “Image Recognition by Combined Affine and
Blur Tchebichef Moment Invariants” 4th International Congress on Image and Signal Processing, (2011)
[38] Ahmad Delforouzi, and M. Pooyan, “Efficient Farsi License Plate Recognition,” 7th International Conference on Information, Communications and Signal Processing (ICICS), Pages 1- 5, (2009)
[39] H. Zhang, W. Jia, X. He, and Q. Wu, “Learning-Based License Plate Detection Using
Global and Local Features,” International Journal of Intelligent Information and Database
Systems (IJIIDS), Vol. 1, No. 2, Pages 228- 243, (2007)
[40] Jitendra Sharma, Amit Sinhal, and Khushboo Saxena, “Comparative Study of Different
Techniques for License Plate Recognition,” JACCT, Vol. 1, No. 2, (2013)
[41] Shreeja Chakraborty, and Ranjan Parekh, “An Improved Template Matching Algorithm
for Car License Plate Recognition,” IJCA, Vol. 118, No. 25, Pages 16- 22, (2015)
[42] Siti Norul Huda Sheikh Abdullah, and Marzuki Khalid, “Comparison of Feature
Extraction in License Plate Recognition,” First Asia International Conference on Modelling,
and Simulation (AMS), Pages 502- 506, (2007)
[43] N. F. Gazcon, and Carlos Ivan Chesnevar, “Automatic Vehicle Identification for
Argentinean License Plates Using Intelligent Template Matching,” National Council of
Scientific and Technical Research (CONICET), Vol. 10, Pages 1066- 1074, (2012)
[44] R. C. Gonzalez, R. E. Woods, and S. L. Eddins, “Digital Image Processing Using MATLAB® ,” 2nd edition, Mc Graw Hill, Singapore, (2011)