照相機便於攜帶，而且成像快速，它相較於掃描器，不僅是更有效率，而且是更為方便的文件影像輸入工具，尤其目前在行動裝置內建照相機的應用越來越普及，例如個人隨身助理 (PDA) 與智慧型手機 (SmartPhone)，使文件影像的處理顯得更為重要。於手持照相機或行動裝置拍攝文件影像時，文件本身的扭曲是很常見的問題，本論文提出一個在不需要其它輔助裝置，只有單一照相機的環境下，復原扭曲文件影像的方法。它可改進現有復原方法的限制，在功能上，可以處理跨頁、雙欄位格式的文件，以及包含圖形或表格的複合文件；在理論上，此方法可不受文件扭曲程度的限制，而且可校正傾斜的文件影像。本論文的主要方法是先分析文件的類型，定位出文件影像中的文字行，接著用立方雲規內插法配適出每條文字行所在的數學函數，然後利用最佳線性組合的原則，挑選出兩條最具代表性的文字行來重建扭曲模型，再予以校正。實驗結果顯示本論文所提的方法能夠有效地校正扭曲和傾斜的文件影像，並且可提高文件影像在後續光學文字辨識(OCR)的正確率。

Cameras are easy to carry and can take images quickly. Compared with a scanner, a camera acting as one of document image inputting devices is not only more efficient but also more portable. Especially at present, the applications of a camera built in the mobile device, like PDAs and Smartphones, are more and more popular, so that the performance of document image processing for such products is getting more important. When we take document images by a hand-held camera or a mobile device, the document suffering from distortion is a common problem. In this thesis, we propose a dewarping method under the environment of capturing images that only needs a single camera without other auxiliary devices. It can improve the restrictions of existing dewaping methods. In terms of function, our method can deal with binding documents, two-column documents, and complex documents which include graphs or tables. On the theoretical plane, our method is not confined to the limit of distortion degrees, and also can rectify tilted document images. The main idea of this thesis is to analyze the document category and locate text lines in the document image. After that, we use the cubic spline method to fit the mathematical functions of the locations of all text lines. Then based on the optimal linear combination principle, we choose two typical text lines to reconstruct the distorted model for dewarping the document image. The experimental results show our method can dewarp the distorted and tilted document images effectively. It also leads to raise the OCR accuracy in the subsequent process of document image analysis.

[1] T. Wada, H. Ukida, and T. Matsuyama, “Shape from shading with interreflections under proximal light source— 3D shape reconstruction of unfolded book surface from a scanner image,” in Proc. of the 15th Int. Conf. on Computer Vision, pp. 66-71, 1995.

[2] T. Araki, H. Guan, and K. Kojima, “Distortion correction technique for scanned book images,” in Ricoh Technical Report, No.29, pp. 31-40, 2003.

[3] Z. Zhang, C. L. Tan and L. Fan, “Estimation of 3D shape of warped document surface for image restoration,” in Proc. of the 17th Int. Conf. on Pattern Recognition, vol. 1, pp.486-489, 2004.

[4] Z. Zhang, C. L. Tan and L. Fan, “Restoration of curved document images through 3D shape modeling,” in Proc. of the 2004 IEEE Computer Society. Conf. on CVPR, vol. 1, pp.10-15, 2004.

[5] Z. Zhang and C. L. Tan, “Restoration of images scanned from thick bound documents,” in Proc. of the 2001 Int. Conf. on Image Processing, vol. 1, pp. 1074-1077, 2001

[6] A. Doncescu, “Former books digital processing: image warping,” in Proc. of the 1997 Workshop on Document Image Analysis, vol. 1, pp. 5-9, 1997.

[7] M. S. Brown and W. B. Seales, “Beyond 2D images: effective 3D imaging for library materials,” in Proc. of the 5th ACM Conf. on Digital Libraries, pp. 27-36, 2000.

[8] M. S. Brown and W. B. Seales, “Document restoration using 3D shape: a general deskewing algorithm for arbitrarily warped documents,” in Proc. of the 8th Int. IEEE Conf. on Computer Vision, vol. 2, pp. 367-374, 2001.

[9] M. S. Brown, W. B. Seales, “Image restoration of arbitrarily warped document,” in IEEE Transaction on Pattern Analysis and Machine Intelligence, vol. 26, pp. 1295-1306, 2004.
[10] M. S. Brown and D. Tsoi, “Correcting common distortions in camera-imaged library materials,” in Proc. of the 2003 Joint Conf. on Digital Libraries, pp. 367-368, 2003.

[11] Y. C. Tsoi and M. S. Brown, “Geometric and shading correction for images of printed materials: a unified approach using boundary,” in Proc. of the 2004 IEEE Computer Society Conf. on CVPR, vol. 1, pp. 240-246, 2004.

[12] H. Cao, X. Ding, and C. Liu, “Rectifying the bound document image captured by the camera: a model based approach,” in Proc. of the 17th Int. Conf. on Document Analysis and Recognition, vol. 1, pp. 71-75, 2003.

[13] H. Cao, X. Ding, and C. Liu, “A cylindrical surface model to rectify the bound document image,” in Proc. of the 9th Int. Conf. on Computer Vision, vol. 1, pp. 228-233, 2003.

[14] Z. Zhang and C. L. Tan, “Straightening warped text lines using polynomial regression,” in Proc. of the 2002 Int. Conf. on Image Processing, vol. 1, pp.977-980, 2002.

[15] Z. Zhang, and C. L. Tan, “Correcting document image warping based on regression of curved text lines,” in Proc. of the 7th Int. Conf. on Document Analysis and Recognition, vol. 1, pp.589-593, 2003.

[16] G. P. Stein, “Lens distortion calibration using point correspondences,” in Proc. of the 16th Int. Conf. on Computer Vision amd Pattern Recognition, pp. 602-608, 1997.

[17] J. Jun and C. Kim, “Robust camera calibration using neural network,” in Proc. of the IEEE Region 10 Conf. on TENCON 99, vol. 1, pp. 694-697, 1999.

[18] B. Tordoff and D. W. Murray, “Violating rotating camera geometry: the effect of radial distortion on self-calibration,” in Proc. of the 15th Int. Conf. on Pattern Recognition, vol. 1, pp. 423-427, 2000.

[19] W. Niblack, An Introduction to Digital Image Processing, Prentice-Hall, Englewood Cliffs, New Jersey, pp. 115-116, 1986.

[20] Z. Zhang and C. L. Tan, “Recovery of distorted document images from bound volumes,” in Proc. of the 16th Int. Conf. on Document Analysis and Recognition, pp. 429-433, 2001.

[21] J. Sauvola, T. Seppanen, S. Haapakoski, and M. Pietikainen, “Adaptive document binarization,” in Proc. of the 4th Int. Conf. on Document Analysis and Recognition, pp. 147-152, 1997.

[22] S. Raju, P. B. Pati, and A. G. Ramakrishnan, “Gabor filter based block energy analysis for text extraction from digital document images,” in Proc. of the 1st Int. Workshop on Document Image Analysis for Libraries, vol. 1, pp. 71-75, 2004.

[23] M. Acharyya, and M. K. Kundu, “Document image segmentation using wavelet scale features,” in IEEE Transactions on Circuits and System for Video Technology, vol. 12, pp. 1117-1127, 2002.

[24] N. Papamarkos, J. Tzortzakis and B. Gatos, “Determination of Run-Length smoothing values for document segmentation,” in Proc. of the 3rd IEEE Int. Conf. On Electronics, Circuits and Systems, ICECS 96, pp. 684-687, 1996.

[25] C. Strouthopoulos, N. Papamarkos, A. Atsalakis, and C. Chamzas, “Locating text in color documents,” in Proc. of 2001 Int. Conf. On Image Processing, vol.1, pp. 1066-1069, 2001.

[26] S. W. Lee and D. S. Ryu, “Parameter-free geometric document layout analysis,” in IEEE Transaction on Pattern Analysis and Machine Intelligence, vol. 23, pp.1240-1256, 2001.

[27] 張任業，應用數值分析，曉圓出版社，1990。

[28] 袁帝文，應用數值方法，儒林圖書公司，1997。

[29] P. F. Hultquist, Numerical Methods, Wendy Calmenson, pp. 144-177, 1988.

[30] R. Sedgewick, Algorithm in C++, Addison Wesley, pp. 545-554, 1992.