不同於以往動態影像車牌辨識較著重於由連續張之糢糊車牌重建清晰車牌，本研究為配合較新式的車牌監視攝影系統，探討由連續張之影像提升車牌偵測、字元切割與辨識效能。本論文主要探討車牌偵測，由於偵測模組在處理單張影像中最為耗時，但結合車牌追蹤能大幅提升車牌偵測在處理動態影像之速度。本研究提出自動學習機制，利用安裝測試時收集的影像擷取車牌運動資訊，以車牌大小、移動軌跡、出現頻率和移動方向等作為學習資訊，經由學習後決定最佳之系統參數，解決車牌必須足夠大小才可偵測之限制，並由此增進即時系統之效能。本研究詳盡探討HOG相關參數對車牌偵測的效能影響，經實驗結論出HOG車牌偵測最佳參數設置。另外探討光流法與Kalman Filter在車牌追蹤的效能，並由此推論不同車牌運動形態下較適合之追蹤方法。經由多樣化的影片樣本證實本系統效能。

The cameras designed for capturing vehicle license plates have been improved in recent years and able to offer sharp image quality at real-time frame rates. Different from many previous works on video-based license plate recognition which mostly focus on transforming multiple low quality images into one clear image good for recognition, this research aims at processing multiple good quality images so that the information extracted from the image sequence can be used to improve plate detection, character segmentation and recognition. The primary focus is on plate detection because it takes a major portion of the processing time when considering the license plate recognition in a single image, and it can be substantially improved when combined with tracking in handling image sequences. This research proposes an automatic learning phase, able to be called upon at system implementation and initial test, that can determine the best operational parameters so that the plate can be detected as soon as it comes into the view of the camera and the characters can be better segmented and recognized. The learning phase would offer the changes in plate size, trajectory, direction and observation frequency to leverage the real-time online performance. A detailed study on the HOG feature parameters good for plate detection is also offered, along with a performance comparison of the optical flow and Kalman filter good for plate tracking. The proposed scheme is tested on various video samples to show its effectiveness.

[1]A. Clemens, F. Limberger, and H. Bischof, "Real-Time License Plate Recognition on an Embedded DSP-Platform," in Computer Vision and Pattern Recognition, 2007. CVPR '07. IEEE Conference on, 2007, pp. 1-8.
[2]M. Donoser, C. Arth, and H. Bischof, "Detecting, tracking and recognizing license plates," presented at the Proceedings of the 8th Asian conference on Computer vision - Volume Part II, Tokyo, Japan, 2007.
[3]I. Giannoukos, C.-N. Anagnostopoulos, V. Loumos, and E. Kayafas, "Operator context scanning to support high segmentation rates for real time license plate recognition," Pattern Recogn., vol. 43, pp. 3866-3878, 2010.
[4]M. S. Sarfraz, A. Shahzad, A. Elahi Muhammad, M. Fraz, I. Zafar, and E. A. Edirisinghe, "Real-time automatic license plate recognition for CCTV forensic applications," Journal of Real-Time Image Processing, pp. 1-11, 2011/11/19 2011.
[5]T. Nicolas, V. Antoine, R. Lionel, and M. Serge, "A cognitive and video-based approach for multinational License Plate Recognition," Machine Vision and Applications, vol. 22, pp. 389-407, 2011/03/01 2011.
[6]H. H. P. Wu, H.-H. Chen, R.-J. Wu, and D.-F. Shen, "License Plate Extraction in Low Resolution Video," in Pattern Recognition, 2006. ICPR 2006. 18th International Conference on, 2006, pp. 824-827.
[7]G.-S. Hsu, J.-C. Chen, and Y.-Z. Chung, "Application-Oriented License Plate Recognition," Vehicular Technology, IEEE Transactions on, vol. 62, pp. 552-561, 2013.
[8]陳明宏, "基於MSER之車牌字元切割和階層式分類器之字元辨識," 台灣科技大學碩士學位論文, 2011.
[9]C. N. E. Anagnostopoulos, I. E. Anagnostopoulos, I. D. Psoroulas, V. Loumos, and E. Kayafas, "License Plate Recognition From Still Images and Video Sequences: A Survey," Intelligent Transportation Systems, IEEE Transactions on, vol. 9, pp. 377-391, 2008.
[10]C. N. E. Anagnostopoulos, I. E. 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, 2006.
[11]K. Zheng, Y. Zhao, J. Gu, and Q. Hu, "License plate detection using Haar-like features and histogram of oriented gradients," in Industrial Electronics (ISIE), 2012 IEEE International Symposium on, 2012, pp. 1502-1505.
[12]J. Yuan, S.-d. Du, and X. Zhu, "Fast super-resolution for license plate image reconstruction," in Pattern Recognition, 2008. ICPR 2008. 19th International Conference on, 2008, pp. 1-4.
[13]K. V. Suresh, G. M. Kumar, and A. N. Rajagopalan, "Superresolution of License Plates in Real Traffic Videos," Intelligent Transportation Systems, IEEE Transactions on, vol. 8, pp. 321-331, 2007.
[14]W. Zeng and X. Lu, "A Generalized DAMRF Image Modeling for Superresolution of License Plates," Intelligent Transportation Systems, IEEE Transactions on, vol. 13, pp. 828-837, 2012.
[15]N. Dalal and B. Triggs, "Histograms of oriented gradients for human detection," in Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on, 2005, pp. 886-893 vol. 1.
[16]J. Matas, O. Chum, M. Urban, and T. Pajdla, "Robust wide baseline stereo from maximally stable extremal regions," Image and Vision, vol. 22, pp. 761-767, 2004.
[17]R. E. Kalman, "A new approach to linear filtering and prediction problems," In Transactions of the ASME - Journal of Basic Engineering, vol. 82, pp. 35–45, 1960.
[18]B. K. P. Horn and B. G. Schunck, "Determining Optical Flow," ARTIFICAL INTELLIGENCE, vol. 17, pp. 185-203, 1981.
[19]B. D. Lucas and T. Kanade, "An iterative image registration technique with an application to stereo vision," presented at the International Joint Conference on Artifical Intelligence, 1981.
[20]J.-Y. Bouguet, "Pyramidal Implementation of the Lucas Kanade Feature Tracker Description of the algorithm," ed: Intel Corporation Microprocessor Research Labs, 2000.
[21]Z. Kalal, K. Mikolajczyk, and J. Matas, "Forward-Backward Error: Automatic Detection of Tracking Failures," in Pattern Recognition (ICPR), 2010 20th International Conference on, 2010, pp. 2756-2759.