隨著科技進步，觸碰面板廣泛地應用於各項數位產品，如：智慧型手機、e-Book、PDA等消費性電子商品，皆可見手寫辨識的應用，因此，本論文主要目的為實現數字與小寫英文之即時手寫辨識系統並驗證其結果。論文內容主要分成硬體架構設計與軟體演算法兩部分：在硬體架構設計部分，包含觸碰面板、微處理器與類比數位轉換器。當觸碰面板表面受到按壓後，觸碰面板獲取類比訊號，經由類比數位轉換器將數據轉至MCU計算座標位置，最後透過RS232傳輸至PC端顯示座標點位置。軟體演算法部分，本系統利用 Visual Studio 2008開發使用者介面平台，採用三層倒傳遞類神經網路作為辨識架構，並分別收集壹仟筆阿拉伯數字及貳仟陸佰筆小寫英文字作為學習樣本，並透過九個同學協助系統測試，針對不同的筆劃進行特徵比對。在市面上，我們可發現手寫辨識產品通常具備著多重候選功能，因此，本系統設計並加入候選字選項的功能，以避免特殊因素造成的辨識失敗。經由測試之證實，本研究提出之手寫辨識系統可有效的辨識出使用者所輸入的字型，其辨識率在合理條件下可達到九成左右，符合市場之一般需求。

With the development of technology, touch panel is widely applied in ubiquitous computing such as smart phone, e-Book, personal digital assistant (PDA). The characteristic capability of these products is the handwriting recognition function. Thus, this thesis contributes to propose a real-time handwriting recognition system with Arabic numbers and lowercase letters, and verify the result with practical tests. This thesis is divided into two parts which are hardware design and software algorithm. In hardware design, the organization contains touch panel, microcontroller and A/D converter. After pressing the touch panel surface, analog signals are obtained and transformed into digital ones by A/D converter. Finally, PC can display the coordinate position through RS232. In software algorithm, the system develops the Graphical User Interface with the software “Visual Studio 2008 C++”. Recognition architecture is constructed by three level back-propagation neural network and learning samples of Arabic numbers and lowercase letters are collected from nine schoolmate. Then, handwriting recognition system is used by different strokes to test its rate of recognition. Because many manufactures of the handwriting recognition system have the function to choice candidate words, this thesis designs the function of candidate words to avoid that some factors make incorrect result. Based on the illustration, the proposed handwriting recognition system of this thesis can achieve about 90% correction rates and can achieve the market standard.

[1] Q. Abbas, J. Ahmad, and W.H. Bangyal, “Analysis of learning rate using BP algorithm for hand written digit recognition application”, Proceeding of International Conference on Information and Emerging Technologies, pp. 1-5, June 2010.
[2] A. Choudhary, R. Rishi, S. Ahlawat and V. S. Dhaka, “Performance analysis of feed forward MLP with various activation functions for handwritten numerals recognition”, Proceeding of the 2nd International Conference on Computer and Automation Engineering, Vol. 5, pp.852-856, February 2010.
[3] R.A. Dilruba, N. Chowdhury, F.F. Liza and C.K. Karmakar, “Data Pattern Recognition with Back-Propagation Training”, Proc. of the 4th International Conference on Electrical and Computer Engineering, pp.451-455, December 2006.
[4] “FS7805 Memory Map & Register Description”, Preliminary, Revision 1.0, 2006.
[5] “FS7805 Security Built-In High Speed USB Peripheral”, FameG, Revision 1.2, 2007.
[6] “Resistive Touch Panel Technologies”, HITACHI Application Note, 2004.
[7] H.J. Lin and S.H. Yen, “A Scheme of on-line Chinese character recognition using neural networks”, Proc. of the IEEE International Conference on Computational Cybernetics and Simulation, Vol. 4, pp.3528-3533, October 1997.
[8] J.E. Dayhoff, “Neural network architectures: an introduction”, New York, N.Y. :Van Nostrand Reinhold, 1990.
[9] M. Chester, “NEURAL NETWORKS A Tutorial”, Englewood Cliffs, N.J. :PTR Prentice Hall, c1993.
[10] M.A. Mazidi, J.G. Mazidi and R.D. McKinlay, “The 8051 microcontroller and embedded systems: using Assembly and C (2nd ed.)”, Upper Saddle River, N.J. :Pearson/Prentice Hall, 2006.
[11] “T5B Chip Design Target Specification”, Ethan Jang, 2007.
[12] “Touch Screen Controller-ADS7846 Datasheet”, Burr-Brown Products from Texas Instruments, 2005.
[13] W. Fang and T. Chang, “Calibration in touch-screen systems”, Analog Applications Journal, Texas Instruments Incorporated-Data Acquisition, pp.5-9, 2007.
[14] X. Wang, T.L. Huang and X.Y. Liu, “Handwritten Character Recognition Based on BP Neural Network”, Third International Conference on Genetic and Evolutionary Computing, pp.520-522, October 2009.
[15] Y. Gao, C. Deng and G. Jiang, “Application of BP neural network in the digital recognition system”, Proc. of the 2nd International Conference on Computer Engineering and Technology, Vol. 6, pp.522-526, April 2010.
[16] Y.W. Bai and C.Y. Chen, “Using Serial Resistors to Reduce the Power Consumption of Resistive Touch Panels”, Proceedings of the 11th Annual IEEE International Symposium on Consumer Electronics, pp. 1-6, June 20-23 2007.
[17] X. Zhang and L. Wu, “Handwritten Digit Recognition Based on Improved Learning Rate BP Algorithm”, Proceedings of the 2th International Conference on Information Engineering and Computer Science, pp. 1-4, Dec. 2010.
[18] Z. Chen and Z. Wei, “A Handwriting Numeral Character Recognition System”, Proc. of International Conference on Multimedia Technology, pp. 1, October 2010.
[19] 王昭智，熱影像技術應用於人臉辨識之研究，國立成功大學工程科學研究所碩士論文，2001年。
[20] 林瑞堂，指尖手寫軌跡的字形辨識，崑山科技大學電機工程研究所碩士論文，2008年。
[21] 郭葵蘭，手寫身分証字號辨識系統，國立高雄第一科技大學電腦與通訊工程研究所，2002年。
[22] 黃詠昌，具數字辨識之3維空間數位手套無線控制器設計與實作，國立成功大學工程科學研究所碩士論文，2009年。
[23] 葉怡成，類神經網路模式應用與實作，儒林圖書有限公司，台灣， 2003年。