This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Supervised Template Estimation for Document Image Decoding
December 1997 (vol. 19 no. 12)
pp. 1313-1324

Abstract—An approach to supervised training of character templates from page images and unaligned transcriptions is proposed. The template training problem is formulated as one of constrained maximum likelihood parameter estimation within the document image decoding framework. This leads to a three-phase iterative training algorithm consisting of transcription alignment, aligned template estimation (ATE), and channel estimation steps. The maximum likelihood ATE problem is shown to be NP-complete and, thus, an approximate solution approach is developed. An evaluation of the training procedure in a document-specific decoding task, using the University of Washington UW-II database of scanned technical journal articles, is described.

[1] Adobe Systems Inc., PostScript Language Reference Manual, 2nd ed. Reading, Mass.: Addison-Wesley, 1990.
[2] H. Baird and G. Nagy, "A Self-Correcting 100-Font Classifier," L. Vincent and T. Pavlidis, eds., Document Recognition: Proc. SPIE, vol. 2,181, pp. 106-115, 1994.
[3] California Dept. of Water Resources, General Comparison of Water District Acts, Bulletin 155-94, Mar. 1994.
[4] F. Chen, D. Bloomberg, and L. Wilcox, "Spotting Phrases in Lines of Imaged Text," L. Vincent and H. Baird eds., Document Recognition II: Proc. SPIE, vol. 2,422, pp. 256-269, 1995.
[5] T. Fruchterman, "DAFS: A Standard for Document and Image Understanding," Proc. 1995 Symp. Document Image Understanding Technology, pp. 4-100,Bowie, Md., Oct.24-25 1995.
[6] J. Hopcroft and J. Ullman, Introduction to Automata Theory, Languages and Computation.Reading, Mass.: Addison-Wesley, 1979.
[7] J. Hull, "Recognition of Mathematics Using a Two-Dimensional Trainable Context-Free Grammar," MEng thesis, Massachusetts Inst. Tech nology, Cambridge, Mass., June 1996.
[8] A. Kam and G. Kopec, "Separable Source Models for Document Image Decoding," L. Vincent and H. Baird, eds., Document Recognition II: Proc. SPIE, vol. 2,422, pp. 84-97, 1995.
[9] A. Kam and G. Kopec, "Document Image Decoding by Heuristic Search," IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 9, pp. 945-950, Sept. 1996.
[10] G. Kopec and P. Chou, "Automatic Generation of Custom Document Image Decoders," Proc. Second Int'l Conf. Document Analysis and Recognition,Tsukuba Science City, Japan, Oct.20-22 1993.
[11] G.E. Kopec and P.A. Chou, “Document Image Decoding Using Markov Source Models,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 16, no. 6, pp. 602-617, June 1994.
[12] G. Kopec and M. Lomelin, "Document-Specific Character Template Estimation," L. Vincent and J. Hull, eds., Document Recognition III: Proc. SPIE, vol. 2,660, pp. 14-26, 1996.
[13] G. Kopec, "Document Image Decoding in the Berkeley Digital Library Project," L. Vincent and J. Hull, eds., Document Recognition III: Proc. SPIE, vol. 2,660, pp. 2-13, 1996.
[14] G. Kopec, "Multilevel Character Templates for Document Image Decoding," L. Vincent and J. Hull, eds., Document Recognition IV: Proc. SPIE, vol. 3,027, 1997.
[15] S.S. Kuo and O. Agazzi, “Keyword Spotting in Poorly Printed Documents Using Pseudo 2-D Hidden Markov Models,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 16, no. 8, pp. 842-848, Aug. 1994.
[16] C. Papadimitriou and K. Steiglitz, Combinatorial Optimization.Englewood Cliffs, N.J.: Prentice Hall, 1982.
[17] I. Phillips, S. Chen, J. Ha, and R. Haralick, Reference Manual for the UW English/Japanese Document Image Database II, Version 2.01, ISL report, Dept. of Electrical Eng., Univ. of Washington, Seattle, Mar.8 1995.
[18] L. Rabiner and B.-H. Juang, Fundamentals of Speech Recognition.Englewood Cliffs, N.J.: Prentice Hall, 1993.
[19] R. Rubenstein, Digital Typography.Reading, Mass.: Addison-Wesley, 1988.
[20] H. Stabler, "Experiences With High-Volume, High-Accuracy Document Capture," L. Spitz and A. Dengel, eds., Document Analysis Systems.Singapore: World Scientific Publishing, 1995.

Index Terms:
Document image decoding, Markov models, template estimation, character recognition, document recognition, maximum likelihood.
Citation:
Gary E. Kopec, Mauricio Lomelin, "Supervised Template Estimation for Document Image Decoding," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, no. 12, pp. 1313-1324, Dec. 1997, doi:10.1109/34.643891
Usage of this product signifies your acceptance of the Terms of Use.