Abstract
Gujarati is an Indic script similar in appearance to other Indo-Aryan scripts. Printed Gujarati script has a rich literary heritage. From an OCR perspective it needs a different treatment due to some of its peculiarities. Research on Gujarati OCR is a recent development as compared to OCR research on many other Indic scripts. Here, in this chapter we present a detailed account of the state of the art of Gujarati document analysis and character recognition. We begin with approaches to zone boundary detection, necessary for the isolation of words and character segmentation and recognition. We show results of various feature extraction techniques such as fringe maps, discrete cosine transform, and wavelets. Zone information and aspect ratios are also used for classification. We present recognition results with two types of classifiers, viz., nearest neighbor classifier and artificial neural networks. Results of experiments wherein various combinations of feature extraction methods with classifiers are also presented. We find that general regression neural network with wavelets feature gives best results with significant time saving in training. Since Indic scripts require syllabic reconstruction from OCR components, a procedure for text generation from the recognized glyph sequences and a method for post-processing is also described.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Two conjuncts /ksha/ and /jya/ are treated as if they are basic consonants in Gujarati script
References
Gandhi, M. K.: Hind Swaraj (Indian Homerule). Navjeevan Publishers
Gandhi, M. K.: Satya-na Prayogo–Atmakatha (My Experiments with Truth - Autobiography). Navjeevan Publishers
Dwyer, R.: The poetics of devotion: the Gujarati lyrics of Dayaram. (2000)
Mehta, S. Y., Dholakia, J.: Gujarati Script. Vishwabharat@TDIL April(2004)
Bansal, V., Sinha, R. M. K.: A Complete OCR for Printed Hindi Text in Devanagari Script. Proc. 6th ICDAR (2001)
Pal, U., Chaudhuri, B. B.: Automatic Separation of Machine-Printed and Hand-Written Text Lines. Proc. of 6th ICDAR (2001) 645–648
Pal, U., Belaid, A., Choisy, C.: Touching Numeral Segmentation Using Water Reservoir Concept. Pattern Recognition Letters 24 (2003) 261–272
Chaudhuri, B.B., Pal, U.: An OCR System to Read Two Indian Lanague Scripts : Bangla and Devanagari. Proc. 4th ICDAR (1997) 1011–1015
Negi, A., Bhagvati, C., Krishna, B.: An OCR System for Telugu. Proc. 4th ICDAR (1997) 1110–1114
Breuel, T.: Segmentation of Hand printed Letter Strings using a Dynamic Programming Algorithm. Proc. of 6th ICDAR (2001) 821–826
Antani, S., Agnihotri, L.: Gujarati Character Recognition. Proc. 6th ICDAR (1999) 418–421
Dholakia, J., Negi, A., S., Ramamohan: Zone Identification in Printed Gujarati Text. Proc. ICDAR (2005) 272–276
Ramamohan, S., Yajnik, A.: Gujarati Numeral Recognition Using Wavelets and Neural Network. IICAI (2005) 397–406
Yajnik, A., Rama Mohan, S.: Identification of Gujarati Characters Using Wavelets and Neural Networks. ISTED Conference on Artificial Intelligence and Soft Computing (2006) 150–155
Dholakia, J., Negi, A., Pathak, V. D.: A Novel Approach To Model Zone Separation Problem In Printed Gujarati Text And Its Solution by Application Of Dynamic Programming. Proc. Of International Conference of Advanced Computing and Communication (2007)
Dholakia, J., Yajnik, A., Negi, A.: Wavelet Feature Based Confusion Character Sets for Gujarati Script. ICCIMA (2007)
Bloomberg, D., Minka, T., Popat, K.: Document Image Decoding using Iterated Complete Path Search with Subsampled Heuristic Scoring. Proc. of 6th ICDAR (2001) 344–349
Popat, K.: Decoding of Text Lines in Grayscale Document Images. Proc. Of ICASSP (2001) 1513–1516
Negi, A., Murthy, K. N., Bhagvati, C.: Issues of Document Engineering in Indian Scripts and Telugu as a Case Study. Vivek (2003)
Gonzalez, R. C., Woods, R. E.: Digital Image Processing. Addison-Wesley (1993)
Pujari, A. K., Naidu, D. C., Sreenivasa Rao, M. Jingara, B. C.: An Adaptive Character Recognizer for Telugu Scripts using Multiresolution Analysis and Associative Memory. Image Vision Computing 22(14) (2004) 1221–1227
Bhattacharya, U., Parui, S. K., Sridhar, M., Kimura, F.: Two-stage Recognition of Handwritten Bangla Alphanumeric Characters Using Neural Classifier. Proc. Of IICAI (2005) 1357–1376
Chaudhuri, A. R., Mandal A. K., Chaudhuri B. B.: Page Layout Analysis for Multilingual Indian Document. Proc. of LEC (2002) 24–32
Duda, O., Heart, P., Stork, D.: Pattern Classification. 2nd. edn. J. Wiley (2001)
Haykin, S.: Neural Networks, A Comprehensive Foundation. Pearson Education Asia (2002)
Specht, D. F.: A General Regression Neural Network. IEEE Transactions on Neural Networks 2(6) (1991) 568–576
Kumar, V. B., Ramakrishnan, A. G.: Radial Basis Function and Subspace Approach For Printed Kannada Text Recognition. Proc. Of ICASSP 5 (2004) 321–324
Amrouche, A., Rouvaen, J. M.: Efficient System for Speech Recognition using General Regression Neural Network. International Journal Of Intelligent Technology 1 (2) (2006) 183–189
Amrouche, A., Rouvaen,J. M.: Arabic Isolated Word Recognition Using General Regression Neural Network. Proc. of the 46th IEEE International Midwest Symposium on Circuits and Systems 2 (2003) 689–692.
Huang, L., Huang, X.: Multi Resolution Recognition Of Off line Handwritten Chinese Characters With Wavelet Transform. Proc. 6th ICDAR (2001) 631–634
Pal, U., Chaudhuri, B. B.: Automatic Separation of Machine-Printed and Hand-Written Text Lines Proc. 6th ICDAR (1999) 645–648
Wang, K. Y., Casey, R. G., Wahl, F. M.: Document Analysis System. IBM J. Res. Development 26 (1982) 647–656
Acknowledgment
Most of this work was supported by the grants from the Ministry of Communications and Information Technology, Government of India, under Resource Center for Indian Language Technology Solutions project and Development Of Robust Document Analysis And Recognition System For Printed Indian Scripts project.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag London Limited
About this chapter
Cite this chapter
Dholakia, J., Negi, A., Mohan, S.R. (2009). Progress in Gujarati Document Processing and Character Recognition. In: Govindaraju, V., Setlur, S. (eds) Guide to OCR for Indic Scripts. Advances in Pattern Recognition. Springer, London. https://doi.org/10.1007/978-1-84800-330-9_4
Download citation
DOI: https://doi.org/10.1007/978-1-84800-330-9_4
Published:
Publisher Name: Springer, London
Print ISBN: 978-1-84800-329-3
Online ISBN: 978-1-84800-330-9
eBook Packages: Computer ScienceComputer Science (R0)