近年來停車場林立，道路監控需求大增，國道計程收費也需要取得車牌號碼當輔助機制，讓車牌辨識成為很重要的技術。本研究將車牌辨識技術分為影像前處理、車牌定位、字元分割、和字元辨識。在影像前處理，先將影像做灰階處理，縮小尺寸以加快處理效率，用直方圖均化及去影像雜訊，增加車牌定位的可靠性。在車牌定位採用Prewitt測邊、膨脹、和連通物件標示法以取得候選的車牌位置。在字元分割，先對影像做二值化處理，做傾斜偵測及校正，再做過濾螺絲釘干擾的字元擷取，及預估未擷取到字元的位置，得到所有字元位置。在字元辨識，針對字元做正規化處理，並根據SOM和LVQ兩種類神經網路來做分類。並且對易混淆字元再做確認，以期獲得較好的實驗結果。在實作方面，先用Matlab實作影像處理和類神經網路。最後移植至TI TMS320DM6446 DSP平台，讓車牌辨識做到真正的產品化。

Recently, parking lots were built everywhere, demand for road surveillance, and Taiwan distance-based toll system also needs to obtain a license plate number for the auxiliary mechanism, the License Plate Recognition (LPR) becomes a very important technology. In this study, the LPR technology was divided into image preprocessing, license plate location, character segmentation and character recognition. In the image preprocessing, does grayscale of image, reduces in size in order to improve the processing efficiency, and goes with histogram equalization and noise filter to increase the reliability. In license plate location, using Prewitt edge detector, dilation, and Connected Components Labeling method to get the candidate of license plate location. In character segmentation, makes the image binarization and does tilt detection and correction. Filters the interfered plate screws, does characters capture, and estimates those not found character position, to get all the character positions. In character recognition, forces normalization on characters, and does classification according to SOM and LVQ neural networks. In order to obtain better results, distinguishes parts of ambiguous characters. In the first implementation, practiced image processing and neural networks by using Matlab. Finally, ported the LPR algorithms to TI TMS320DM6446 DSP platform to achieve a real product.

[1] http://www.appledaily.com.tw/appledaily/article/supplement/
20130610/35073745/
[2] http://web1.nsc.gov.tw/ct.aspx?xItem=14263&ctNode=40
[3] C. Anagnostopoulos, I. Anagnostopoulos, E. Kayafas, and V. Loumos,“A license plate recognition system for intelligent transportation system applications,” IEEE Trans. Intell. Transp. Syst., vol. 7, no. 3, pp. 377–392, Sep. 2006.
[4] C. Anagnostopoulos, I. Anagnostopoulos, I. Psoroulas, V. Loumos, and E. Kayafas, “License Plate Recognition From Still Images and Video Sequences: A Survey”, IEEE Trans. Intell. Transp. Syst., vol. 9, no. 3, pp. 377-391, Sep. 2008.
[5] N. Bellas, S. M. Chai, M. Dwyer, and D. Linzmeier, “FPGA implementation of a license plate recognition SoC using automatically generated streaming accelerators,” in Proc. 20th IPDPS, Nice, France, Apr. 2006,pp. 8–15.
[6] P. Wu, H.-H. Chen, R.-J. Wu, and D.-F. Shen, “License plate extraction in low resolution video,” in Proc. 18th ICPR, Hong Kong, 2006, vol. 1,pp. 824–827.
[7] A. Broumandnia and M. Fathy, “Application of pattern recognition forFarsi license plate recognition,” in Proc. Int. Conf. GVIP, Cairo, Egypt,2005.
[8] S. Nomura, K. Yamanaka, O. Katai, H. Kawakami, and T. Shiose,“A novel adaptive morphological approach for degraded character image segmentation,” Pattern Recognition, vol. 38, no. 11, pp. 1961–1975, Nov. 2005.
[9] P. Stec and M. Domanski, “Efficient unassisted video segmentation using enhanced fast marching,” in Proc. ICIP, vol. 3, pp. 427–430.
[10] L. Xu, A. Krzyzak, and C. Y. Suen, “Methods of combining multiple classifiers and their application to handwriting recognition,” IEEE Trans. Syst., Man, Cybern., vol. 22, no. 3, pp. 418–435, May/Jun. 1992.
[11] S.-L. Chang, L.-S. Chen, Y.-C. Chung, and S.-W. Chen, “Automatic license plate recognition,” IEEE Trans. Intell. Transp. Syst., vol. 5, no. 1, pp. 42–53, Mar. 2004.
[12] TMS320DM6446 Digital Media System-on-Chip, Texas Instruments Incorporated, 2010.
[13] DaVinci Technology Overview, Texas Instruments Incorporated, 2008.
[14] Linux Embedded System Design Workshop Student Notes，Designing with Texas Instruments ARM and ARM+DSP Systems, Texas Instruments Incorporated, 2011.
[15] Codec Engine Application Developer User's Guide, Texas Instruments Incorporated, 2009.
[16] http://www.ti.com/tool/tmdsevm6446
[17] http://www.sanhengxing.com/cpdm6446.asp
[18] 賴建庭，「混合專家模型應用於影像車牌辨識」，資訊工程系碩士論文，南台科技大學，台南(1997)。
[19] TMS320DM644x DMSoC Video Processing Front End (VPFE) User's Guide, Section 2.3.2, Texas Instruments Incorporated, 2010.
[20] R. Gonzalez and R. Woods, Digital image Processing, 2nd Edition, Prentice-Hall, 2002.
[21] http://en.wikipedia.org/wiki/Connected-component_labeling
[22] N. Otsu, “A threshold selection method from gray level histograms”, IEEE Trans. on Systems, Man, and Cybernetics, Vol. SMC-9, pp. 62-66, 1979.
[23] 鍾國亮，「影像處理與電腦視覺」，東華書局，台北，第128-139頁（2006）。
[24] 張斐章、張麗秋，「類神經網路」，東華書局，台北，第229-244頁（2006）。
[25] 葉怡成，「類神經網路模式應用與實作」，儒林圖書有限公司，台北，第6.1-6.21頁，第8.1-8.26頁(2006)。
[26] http://www.ti.com/tool/sprc264
[27] http://www.ti.com/tool/sprc831