本研究旨在針對表單文件自動化處理進行研究，針對表單處理中之手寫欄位分類、擷取與手寫資料擷取等問題提出解決的方法。在表單手寫欄位擷取的階段，分別利用表單中物件的尺寸大小、比例、物件整體性結構特性與物件方向性結構特徵，作為物件之分類特徵。為便於取得物件之結構特徵，本研究利用影像編碼的方式，將空白表單影像轉換成簡化的結構圖。同時為區辨說明欄位與包含說明文字之填寫欄位，分別利用欄位區域水平及垂直方向之像素投影，配合說明文字之分佈、大小與文字間距等特徵，進行分析辨識。
在手寫資料擷取的階段中，將已填寫之表單影像與已知空白表單樣本進行比
對後，根據相同類別的空白表單之手寫欄位資訊，擷取出已填寫表單中之手寫欄位資料。對於所擷取出之手寫資料中，因框線去除後，造成與框線相交之手寫筆畫斷裂的問題，提出判斷筆畫相交區段，並重建相交區段之手寫筆畫的方法，修補破碎手寫筆畫。
本研究之測試影像，共分為一般單純格式之表單影像與格式複雜之複合式表
單影像等兩類。由實驗結果可證明本研究所提出之方法，針對不同類型之表單影像，皆可得到不錯的效果。

Form document analysis is one of the most essential tasks in document analysis and recognition. The problems of form fields and filled-in data extraction are two important parts of form document analysis. For form field extraction, the first major task was to classify the preprinted text, lines, check boxes, text boxes and the tables of a form. This thesis proposes a method which based on direction-invariant global structural features and directional dependant structural features to classify the form fields, and then extract the filled-in spaces in a form document. Since tables can contain both name fields and data fields, for the second task, we used a method based on horizontal and vertical color histogram distribution features to segment the fields and extract the data fields. For filled-in data extraction, we propose a method which based on Run-based algorithm and the idea of interpolation to detect the character strokes overlapped by printed form frame and reconstruct the broken strokes after removing the frame line. The experimental results on different types of form
documents showed a 99% recognition rate on form fields extraction, and a 91% successful filled-in data extraction rate was achieved.

表目錄 ..................................................iii
圖目錄 ...................................................iv
第一章 緒論.................................................1
1.1 研究動機與目的..........................................1
1.2 研究範圍與限制..........................................3
1.3 論文結構................................................4第二章 文獻探討.............................................5
2.1 表單文件處理............................................6
2.1.1 結構切割..............................................6
2.1.2 結構表示方式..........................................8
2.1.3 區塊物件識別..........................................8
2.1.4 相似度比對............................................9
2.2 手寫資料處理............................................9
2.2.1 框線去除及破碎字修補...................................9
2.2.2 光學字元辨識.........................................10
第三章 系統簡介............................................11
第四章 方法與步驟...........................................15
4.1 表單物件的分類及結構分析.................................15
4.1.1 表單物件分類.........................................18
4.1.2 物件結構分析.........................................21
4.2 虛線重組...............................................26
4.3 底線去除...............................................29
4.4 填寫欄及說明欄之區別....................................32
4.4.1 表格欄位之擷取.......................................32
4.4.2 表格與方格填寫欄位之擷取...............................33
4.5 框線去除及破碎字修補....................................44
第五章 實驗結果與討論.......................................46
5.1 實驗資料來源...........................................46
5.2 實驗驗證...............................................48
5.2.1 實驗一、簡單結構之問卷式表單文件處理....................49
5.2.2 實驗二、複雜結構之複合式表單文件處理....................53
5.3 總結..................................................60
第六章 結論與未來研究.......................................61
6.1 結論..................................................61
6.2 未來研究...............................................62
參考文獻...................................................63
附錄 實驗用表單文件範本.....................................67

[1] S. Di Zenzo, L. Cinque, and S. Levialdi, “Run-based Algorithm for Binary Image Analysis and Processing,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 1, pp. 83-89, 1996.
[2] R. Zanibbi, D. Blostein, J. R. Cordy, “A Survey of Table Recognition,” International Journal on Document Analysis and Recognition, vol. 7, no. 1, pp. 1-16, 2004.
[3] Y. F. Zheng, C. S. Liu, X. Q. Ding and S. Y. Pan, “Form Frame Line Detection with Directional Single-Connected Chain,” Proc. Int. Conf. Document Analysis and Recognition, pp. 699-703, 2001.
[4] Y. F. Zheng, H. P. Li and D. Doermann, “A Parallel-Line Detection Algorithm Based on HMM Decoding,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 27, no. 5, pp. 777-790, 2005.
[5] H. E. Nielson and W. A. Barrett, “Consensus-Based Table Form Recognition,” Proc. Int. Conf. Document Analysis and Recognition, pp. 906-910, 2003.
[6] Y. Y. Tang, H. Ma, J. M. Liu, B. F. Li and D. H. Xi, “Multiresolution Analysis in Extraction of Reference Lines from Documents with Gray Level Background,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 19, no. 8, pp. 921-926, 1997.
[7] D. H. Xi and S. W. Lee, “Reference Line Extraction from Form Documents with Complicated Backgrounds,” Proc. Int. Conf. Document Analysis and Recognition, pp. 080-084, 2003.
[8] T. M. Lu and K. C. Fan, “Form Segmentation and Recognition by Clustering of Feature Points and Matching of Feature Graphs,” Proc. Workshop of Center of Excellence for Computer System Technology, pp. 120-133, 1994.
[9] S. W. Lam, L. Javanbakht and S. N. Srihari, “Anatomy of A Form Reader,” Proc. Int. Conf. Document Analysis and Recognition, pp. 506-509, 1993.
[10] P. Duygulu and V. Atalay, “A Hierarchical Representation of Form Documents for Identification and Rereival,” International Journal on Document Analysis and
Recognition, vol. 5, no. 1, pp. 17-27, 2002.
[11] J. H. Liu and A. K. Jain, “Image-based Form Document Retrieval,” Proc. Int. Conf. Pattern Recognition, vol. 1, pp. 626-628, 1998.
[12] J. H. Liu, X. Q. Ding and Y. S. Wu, “Description and Recognition of Form and Automated Form Data Entry,” Proc. Int. Conf. Document Analysis and Recognition, pp. 579-582, 1995.
[13] A. Busch, W. W. Boles, S. Sridharan and V. Chandran, “Detection of Unknown Forms from Document Images,” Proc. Workshop on Digital Image Computing, pp. 141-144, 2003.
[14] T. Watanabe and T. Sobue, “Layout Analysis of Complex Documents,” Proc. Int. Conf. Pattern Recognition, vol. 4, pp. 447-450, 2000.
[15] T. Watanabe, Q. Luo and N. Sugie, “Layout Recognition of Multi-Kinds of Table-Form Documnets,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 17, no. 4, pp. 432-445, 1995.
[16] K. Luo, S. latifi, K. Taghva and E. Regentova, “Recognition and Identification of Form Document Layouts,” Proc. Int. Conf. Information Technology: Coding and Computing, pp. 352-356, 2004.
[17] M. Diligenti, P. Frasconi and M. Gori, “Hidden Tree Markov Models for Document Image Classification,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 25, no. 4, pp. 519-523, 2003.
[18] A. Amano and N. Asada, “Graph Grammar Based Analysis System of Complex Table Form Document,” Proc. Int. Conf. Document Analysis and Recognition, vol. 2, pp. 916-920, 2003.
[19] A. Amano, N. Asada and M. Mukunoki, “Modification Table Form Generation System based on The Form Recognition,” Proc. Int. Conf. Document Analysis and Recognition, vol. 2, pp. 659-662, 2004.
[20] A. Amano, N. Asada, T. Motoyama, T. Sumiyoshi and K. Suzuki, “Table Form Document Synthesis by Grammar-Based Structure Analysis,” Proc. Int. Conf. Document Analysis and Recognition, pp. 533-537, 2001.
[21] K. C. Fan, Y. K. Wang and M. L. Chang, “Form Document Identification Using Line Structure Based Features,” Proc. Int. Conf. Document Analysis and Recognition, pp. 704-708, 2001.
[22] H. C. Peng, F. H. Long, W. C. Siu, Z. R. Chi and D. D. G. Feng, “Document Image Matching Based on Component Blocks,” Proc. Int. Conf. Image Processing, pp. 601-604, 2000.
[23] S. Shimotsuji and M. Asano, “Form Identification based on Cell Structure,” Proc. Int. Conf. Pattern Recognition, pp. 793-797, 1996.
[24] 張貴雲，“表單手寫欄位資料之萃取”，碩士論文，國立師範大學資訊教育研究所，民國八十七年六月。
[25] S. Tabbone, L. Wendling and K. Tombre, “Matching of Graphical Symbols in Line-drawing Images Using Angular Signature Information,” Int. Journal on Document Analysis and Recognition, vol. 6, pp. 115-125, 2003.
[26] M. Ye, M. Bern and D. Goldberg, “Document Image Matching and Annotation Lifting,” Proc. Int. Conf. Document Analysis and Recognition, pp. 753-760, 2001.
[27] B. Yu and A. K. Jain, “A Generic System for Form Dropout,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 11, pp. 1127-1134, 1996.
[28] J. S. Chen and D. C. Tseng, “Overlapped-Character Separation and Reconstruction for Table-Form Documents,” Proc. Int. Conf. Image Processing, pp. 233-236, 1996.
[29] G. Nagy, “Twenty Years of Document Image Analysis in PAMI,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 22, no. 1, pp. 38-62, 2000.
[30] 林育慈，“離線手寫印刷體英數字之辨識”，碩士論文，國立師範大學資訊教育研究所，民國八十七年六月。
[31] S. Marinai, M. Gori and G. Soda, “Artificial Neural Networks for Document Analysis and Recognition,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 27, no. 1, pp. 23-35, 2005.
[32] C. F. Liou and H. C. Yang, “Hand-printed Character Recognition based on Spatial Topology Distance Measurement,” IEEE Trans. Pattern Analysis and Machine
Intelligence, vol. 18, no. 9, pp. 941-945, 1996.
[33] D. F. Chang, J. Y. J. Hsu and C. S. Fuh, “Handwritten Character Recognition Using a Neural Network,” Proc. OCR & DA, pp. 17-20, 1996.
[34] H. W. Hao, X. H. Xizo and R. W. Dai, “Handwritten Chinese Character Recognition by Metasynthetic Approach,” Pattern Recognition, vol. 30, no. 8, pp. 1321-1328, 1997.
[35] H. H. Tseng and W. H. Tsai, “Character Font Recognition Using Fourier Spectrum Features and Back-propagation Neural Network,” Proc. Int. Conf. Computer
Vision Graph and Image Processing, pp. 206-213, 1996.
[36] K. C. Fan and L. S. Wang, “A Run Length Histogram Based Approach to the Identification of Machine-printed and Handwritten Chinese Text Images,” Proc. Int. Conf. Computer Vision Graph and Image Processing, pp. 416-419, 1996.
[37] H. Shinjo, E. Handano, K. Marukawa, Y. Shima and H. Sako, “A Recursive Analysis for Form Cell Recognition,” Proc. Int. Conf. Document Analysis and Recognition, pp. 694-698, 2001.