隨著LCD產業在次世代的競爭對手產能更大、而本業人力費用更加高昂的情況下，除了強化企業技術能力及拓展銷售市場為必要手段之外，如何在每個生產環節撙節成本，成為LCD產業能在一片藍海中殺出一條血路的重要標的。
以LCD前段製程為例，從原料玻璃投入，經鍍膜、塗佈、曝光、顯影、蝕刻、檢測、注入液晶、重合、切割到偏光板貼附後產品產出等，可謂是自動化程度相當高的產業。其中玻璃為LCD產品中成本最高的原料，在數百道的自動化生產過程中，由刻號機雷射刻蝕的玻璃ID為自動化驗證帳料是否正確的重要索引值，如刻號過程因玻璃定位失準造成ID刻號錯誤則可能導致接下來的製程參數異常、等級分類錯誤及出貨異常等損失。正確ID的重要性可見一斑。而在精實生產的條件下並無法投入大量人力驗證ID刻號的準確性，所以必須找出其它方法來代替人力檢查每一個玻璃ID。故研究以字元辨識方式自動辨識ID刻號的準確性以取代人力檢查。
本研究提出一套ID刻號發生偏移的情況下仍能實現高辨識率的演算流程，由測試結果分析來看，本篇論文之演算流程的確有效地降低複雜度且於刻號偏移後字元剩餘50%以上的情況下仍保有高辨識率的優點，可謂是相當實用之系統。
而將本研究導入產業的具體成果方面，本系統可節省ID檢驗的人力，以往由人工抽檢的方式，導入本系統可提高抽檢率至100%並將人力需求降至0%，同時可立即回報ID刻號異常結果以減少更多不必要的刻號異常情況發生，而降低重新刻號的發生率意謂著生產成本及不良率也能隨之降低。

When the LCD industry marching into new generations, companies have been confronting competitors with larger production capacity and more expensive manpower costs. While strengthening the core technical capabilities of an enterprise and expanding the market sales are the necessary means, one way the LCD industry could fight its way out would be reducing the production costs in every aspect and it has become an important indicator in the red ocean.

The essence of the LCD industry is the realization of automated production technology. In the LCD automated manufacturing, after glass raw materials coming to the production line and going through plating, etching, exposure, development, testing, liquid crystal injection, combination, panel cutting, polarizing plate attaching, and product shipments, Several hundred automated production processes need to be completed. Since glass substrates are the most critical part and are the most expensive raw materials in the LCD products. During the course of the automated production, glass ID engraved by laser etching machine is an important index for production traceability, identification, and inquiries. Engraved glass ID number may be incomplete or inaccurate due to the erroneous etching position on the glass and may result in abnormal productions, error product grade classifications, and resume shipments could not be retroactive and other conditions, so the importance of proper ID engraved evident.

However, under highly automated manufacturing conditions, it is not practical to put a lot of manpower for verifying the accuracy of the ID numbers engraved manually. Companies must find other ways to inspect every glass ID efficiently and correctly. This study proposes and implements an automatic character recognition method for identifying engraved glass ID numbers to replace manpower inspection.

[1] Draghici, S.(1997), A neural network based artificial vision system for license plate recognition, Int. Journal of Neural Systems, vol.8, no.1, pp.113-126, Feb. 1997.
[2] Kamat, V. and S. Ganesan(1995), An efficient implementation of the Hough transform for detecing vehicle license plates using DSP'S, in Proc. Conf. Real-Time Technology and Application., Chicago, IL, May 15-17, 1995, pp.58-59.
[3] Lee, H.-J., S.-Y. Chen, and S.-Z. Wang(2004), Extraction and recognition of license plates of motocycles and vehicles on highways, in Proc. 17th Int. Conf. Pattern Recognition, Hsinchu, Taiwan, Aug.23-26, 2004, pp.356-359.
[4] 黃錦溢(2000)，車牌自動辨識系統之設計，淡江大學電機工程學系碩士論文。
[5] 陳克智(2011)，照相手機的車牌偵測與辨識，國立中央大學資訊工程研究所碩士論文。
[6] Y. Zhang and C. Zhang(2003), A New Algorithm for Character Segmentation of License Plate, in Proc, of IEEE Int. Conf, on Intelligent Vehicles Symposium, pp. 106-109, June 2003.
[7] Y. P. Huang, S. Y. Lai, and W. P. Chuang(2004), A Template-Based Model for License Plate Recognition, in Proc. IEEE Int. Conf. on Networking, Sensing and Control, vol. 2, pp. 737-742, March 2004.
[8] H.A. Hegt, R.J Haye, and N.A. Khan(1998), A High Performance License Plate Recognition System, Systems, Man, and Cybernetics, 1998. 1998 IEEE International Conference on Volume: 5 , 1998 , Page(s): 4357 -4362 vol.5.
[9] 溫福助(2000)，類神經網路樣板比對法於車牌字元辨識之研究，台灣大學電機工程學研究所碩士論文。
[10] 張勝仁(2003)，一個可增強車牌字元辨識效能的多重辨識器設計，中原大學電子工程系碩士論文。
[11] Comelli, P., P. Ferragina, M. N. Granieri(1995), and F. Stabile, Optical recognition of motor vehicle license plates, IEEE Trans. Vehicular Technology. vol.44, no.4, pp.790-799, Nov. 1995.
[12] J. Barroso, J. Bulas-Cruz, E.L. Dagless, and A. Rafael(1997), Number Plate Reading Using Computer Vision, Industrial Electronics, ISIE '97., Proceedings of the IEEE international Symposium on , 1997 , Page(s): 761 –766 vol.3.
[13] Pandya, A. S., Macy, R. B. (1996). Pattern Recognition with Neural Networks in C++, CRC PRESS & IEEE PRESS.
[14] Lu, Y.(1993), On the segmentation of touching characters, in Proc. 2nd Int. Conf. on Document Analysis and Recognition, Tsukuba Science City, Japan, Oct.20-22, 1993, pp.440-443.
[15] 莊志鴻(1999)，以 WMMS 法做車牌字元辨識，國立交通大學電機語控制工程學系碩士論文。
[16] N. Otsu(1979), A Thresholding Selection Method from Gray Level Histograms, IEEE Transactions on Systems, Man and Cybernetics , vol. 9, pp. 62-66.