Abstract
The Arabic alphabet is used in around 27 languages, including Arabic, Persian, Kurdish, Urdu, and Jawi. Many researchers have developed systems for recognizing cursive handwritten Arabic words, using both holistic and segmentation-based approaches. This paper introduces a system that achieves high accuracy using efficient segmentation, feature extraction, and recurrent neural network (RNN). We describe a robust rule-based segmentation algorithm that uses special feature points identified in the word skeleton to segment the cursive words into graphemes. We show that careful selection from a wide range of features extracted during and after the segmentation stage produces a feature set that significantly reduces the label error. We demonstrate that using same RNN recognition engine, the segmentation approach with efficient feature extraction gives better results than a holistic approach that extracts features from raw pixels. We evaluated this segmentation approach against an improved version of the holistic system MDLSTM that won the ICDAR 2009 Arabic handwritten word recognition competition. On the IfN/ENIT database of handwritten Arabic words, the segmentation approach reduces the average label error by 18.5 %, the sequence error by 22.3 %, and the execution time by 31 %, relative to MDLSTM. This approach also has the best published accuracies on two IfN/ENIT test sets.
Similar content being viewed by others
References
Abandah, G., Jamour, F.: Recognizing handwritten Arabic script through efficient skeleton-based grapheme segmentation algorithm. In: International Conference Intelligent Systems Design and Applications, pp. 977–982 (2010)
Abandah, G., Jamour, F.: A word matching algorithm in handwritten Arabic recognition using multiple-sequence weighted edit distances (2014). Submitted.
Abandah, G., Khedher, M.: Analysis of handwritten Arabic letters using selected feature extraction techniques. Int. J. Comput. Process. Lang. 22(1), 49–73 (2009)
Abandah, G., Malas, T.: Feature selection for recognizing handwritten Arabic letters. Dirasat Eng. Sci. J. 37(2), 242–256 (2010)
Al-Hajj, R., Likforman-Sulem, L., Mokbel, C.: Arabic handwriting recognition using baseline dependant features and hidden Markov modeling. In: International Conference Document Analysis and Recognition, pp. 893–897 (2005)
Alginahi, Y.M.: A survey on Arabic character segmentation. Int. J. Doc. Anal. Recogn. 16(2), 105–126 (2013)
Alkhoury, I., Giménez, A., Juan, A.: Arabic handwriting recognition using Bernoulli HMMs. In: Märgner, V., El Abed, H. (eds.) Guide to OCR for Arabic Scripts, pp. 255–272. Springer, London (2012)
Amin, A.: Arabic character recognition. In: Bunke, H., Wang, P. (eds.) Handbook of Character Recognition and Document Image Analysis, pp. 397–420. World Scientific, Singapore (1997)
Arica, N., Yarman-Vural, F.: Optical character recognition for cursive handwriting. IEEE Trans. Pattern Anal. Mach. Intell. 24(6), 801–813 (2002)
Azeem, S.A., Ahmed, H.: Effective technique for the recognition of offline Arabic handwritten words using hidden Markov models. Int. J. Doc. Anal. Recogn. 16(4), 399–412 (2013)
Chang, F., Chen, C.J., Lu, C.J.: A linear-time component-labeling algorithm using contour tracing technique. Comput. Vis. Image Underst. 93(2), 206–220 (2004)
Deutsch, E.: Thinning algorithms on rectangular, hexagonal, and triangular arrays. Commun. ACM 15(9), 827–837 (1972)
Douglas, D., Peucker, T.: Algorithms for the reduction of the number of points required to represent a line or its caricature. Canad. Cartogr. 10(2), 112–122 (1973)
Dreuw, P., Rybach, D., Heigold, G., Ney, H.: RWTH OCR: A large vocabulary optical character recognition system for Arabic scripts. In: Märgner, V., El Abed, H. (eds.) Guide to OCR for Arabic Scripts, pp. 215–254. Springer, London (2012)
El Abed, H., Märgner, V.: Comparison of different preprocessing and feature extraction methods for offline recognition of handwritten Arabic words. In: International Conference Document Analysis and Recognition, pp. 974–978 (2007)
El-Hajj, R., Likforman-Sulem, L., Mokbel, C.: Arabic handwriting recognition using baseline dependant features and hidden Markov modeling. In: International Conference Document Analysis and Recognition, pp. 893–897 (2005)
Freeman, H.: On the encoding of arbitrary geometric configurations. IRE Trans. Electron. Comput. 10(2), 260–268 (1961)
Gers, F.: Long short-term memory in recurrent neural networks. Ph.D. thesis, Ecole Polytechnique Fédérale de Lausanne (2001)
Graves, A.: RNNLIB: a recurrent neural network library for sequence learning problems. http://sourceforge.net/projects/rnnl/
Graves, A.: Supervised sequence labelling with recurrent neural networks. Ph.D. thesis, Technische Universität München (2008)
Graves, A.: Offline Arabic handwriting recognition with multidimensional recurrent neural networks. In: Märgner, V., El Abed, H. (eds.) Guide to OCR for Arabic Scripts, pp. 297–313. Springer, London (2012)
Graves, A.: Supervised Sequence Labelling with Recurrent Neural Networks. Studies in Computational Intelligence, vol. 382. Springer, Berlin (2012)
Graves, A., Fernández, S., Gomez, F., Schmidhuber, J.: Connectionist temporal classification: Labelling unsegmented sequence data with recurrent neural networks. In: International Conference Machine Learning (2006)
Graves, A., Fernández, S., Liwicki, M., Bunke, H., Schmidhuber, J.: Unconstrained online handwriting recognition with recurrent neural networks. Adv. Neural Inf. Process. Syst. 20, 1–8 (2008)
Graves, A., Fernández, S., Schmidhuber, J.: Multi-dimensional recurrent neural networks. In: International Conference Artificial Neural Networks (2007)
Graves, A., Liwicki, M., Fernández, S., Bertolami, R., Bunke, H., Schmidhuber, J.: A novel connectionist system for unconstrained handwriting recognition. IEEE Trans. Pattern Anal. Mach. Intell. 31(5), 855–868 (2009)
Graves, A., Schmidhuber, J.: Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Netw. 18(5), 602–610 (2005)
Graves, A., Schmidhuber, J.: Offline Handwriting Recognition with Multidimensional Recurrent Neural Networks. In: Advances in Neural Information Processing Systems, NIPS’22, vol. 22, chap., pp. 545–552. MIT Press, Vancouver (2009)
Guyon, I., Elisseeff, A.: An introduction to variable and feature selection. J. Mach. Learn. Res. 3(1), 1157–1182 (2003)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997)
Kuhl, F., Giardina, C.: Elliptic Fourier features of a closed contour. Comput. Graph. Image Process. 18(3), 236–258 (1982)
Kundu, A., Hines, T., Phillips, J., Huyck, B.D., Van Guilder, L.C.: Arabic handwriting recognition using variable duration HMM. In: International Conference Document Analysis and Recognition, pp. 644–648 (2007)
Lee, H., Verma, B.: Binary segmentation algorithm for English cursive handwriting recognition. Pattern Recogn. 45(4), 1306–1317 (2012)
Lewis, M.P. (ed.): Ethnologue: Languages of the World. SIL International, Dallas (2009)
Likforman-Sulem, L., Mohammad, R.A.H., Mokbel, C., Menasri, F., Bianne-Bernard, A.L., Kermorvant, C.: Features for HMM-based Arabic handwritten word recognition systems. In: Märgner, V., El Abed, H. (eds.) Guide to OCR for Arabic Scripts, pp. 123–143. Springer, London (2012)
Liu, C.L.: Handwritten Chinese character recognition: effects of shape normalization and feature extraction. In: Doermann, D., Jaeger, S. (eds.) Arabic and Chinese Handwriting Recognition, vol. LNCS 4768, pp. 104–128. Springer, Berlin (2008)
Lorigo, L., Govindaraju, V.: Segmentation and pre-recognition of Arabic handwriting. In: International Conference Document Analysis and Recognition, pp. 605–609 (2005)
Lorigo, L., Govindaraju, V.: Offline Arabic handwriting recognition: a survey. IEEE Trans. Pattern Anal. Mach. Intell. 28(5), 712–724 (2006)
Maddouri, S.S., El-Abed, H., Samoud, F.B., Bouriel, K., Ellouze, N.: Baseline extraction: comparison of six methods on IFN/ENIT database. In: International Conference Frontiers in Handwriting Recognition (2008)
Märgner, V., El Abed, H.: ICDAR 2005—Arabic handwriting recognition competition. In: International Conference Document Analysis and Recognition, pp. 70–74 (2005)
Märgner, V., El Abed, H.: ICDAR 2007—Arabic handwriting recognition competition. In: International Conference Document Analysis and Recognition, pp. 1274–1278 (2007)
Märgner, V., El Abed, H.: ICDAR 2009—Arabic handwriting recognition competition. In: International Conference Document Analysis and Recognition, pp. 1383–1387 (2009)
Märgner, V., El Abed, H.: ICFHR 2010—Arabic handwriting recognition competition. In: International Conference Frontiers in Handwriting Recognition, pp. 709–714 (2010)
Märgner, V., El Abed, H.: ICDAR 2011—Arabic handwriting recognition competition. In: International Conference Document Analysis and Recognition, pp. 1444–1448 (2011)
Motawa, D., Amin, A., Sabourin, R.: Segmentation of Arabic cursive script. In: International Conference Document Analysis and Recognition, pp. 625–628 (1997)
Murray, A., Edwards, P.: Synaptic weight noise during multilayer perceptron training: fault tolerance and training improvements. IEEE Trans. Neural Netw. 4(4), 722–725 (1993)
Pechwitz, M., El Abed, H., Märgner, V.: Handwritten Arabic word recognition using the IFN/ENIT-database. In: Märgner, V., El Abed, H. (eds.) Guide to OCR for Arabic Scripts, pp. 169–213. Springer, London (2012)
Pechwitz, M., Maddouri, S.S., Märgner, V., Ellouze, N., Amiri, H.: IFN/ENIT - database of handwritten Arabic words. In: Colloque Int’l Francophone sur l’Ecrit et le Document, pp. 129–136 (2002)
Peng, H.: mRMR (minimum redundancy maximum relevance feature selection). http://penglab.janelia.org/proj/mRMR/
Peng, H., Long, F., Ding, C.: Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans. Pattern Anal. Mach. Intell. 27(8), 1226–1238 (2005)
Qaralleh, M., Abandah, G., Jamour, F.: Tuning recurrent neural networks for recognizing handwritten Arabic words. J. Softw. Eng. Appl. 6(10), 533–542 (2013)
Ratcliff, J., Metzener, D.: Pattern matching: the Gestalt approach. Dr. Dobb’s J. 13(7), 46–72 (1988)
Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning internal representations by error propagation. In: Rumelhart, D.E., McClelland, J.L. (eds.) Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol. 1: Foundations, Chap., pp. 318–362. MIT, Cambridge (1986)
Safabakhsh, R., Adibi, P.: Nastaaligh handwritten word recognition using a continuous-density variable-duration HMM. Arab. J. Sci. Eng. 30(1B), 95–118 (2005)
Sari, T., Souici, L., Sellami, M.: Off-line handwritten Arabic character segmentation algorithm: ACSA. In: International Workshop on Frontiers in Handwriting Recognition, pp. 452–457 (2002)
Schambach, M.P., Rottland, J., Alary, T.: How to convert a Latin handwriting recognition system to Arabic. In: International Conference Frontiers in Handwriting Recognition (2008)
Schuster, M., Paliwal, K.K.: Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 45, 2673–2681 (1997)
Smith, R.: An overview of the Tesseract OCR engine. In: International Conference Document Analysis and Recognition, pp. 629–633 (2007)
Wshah, S., Shi, Z., Govindaraju, V.: Segmentation of Arabic handwriting based on both contour and skeleton segmentation. In: International Conference Document Analysis and Recognition, pp. 793–797 (2009)
Acknowledgments
This work was supported by the Deanship of the Scientific Research in the University of Jordan. Some of this research was completed when G. Abandah was in a sabbatical leave in Princess Sumaya University for Technology. We would like to thank Alex Graves for making the RNNLIB publically available [19], for giving us a copy of the latest RNNLIB version, and for his help in using it. We also thank him for providing parts of the RNN sequence transcriber description included in Sect. 4. We would also like to thank Hanchuan Peng for making mRMR tools publically available [49]. We would like also to thank Haikal El Abed for giving us copies of sets f and s of the IfN/ENIT database.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Abandah, G.A., Jamour, F.T. & Qaralleh, E.A. Recognizing handwritten Arabic words using grapheme segmentation and recurrent neural networks. IJDAR 17, 275–291 (2014). https://doi.org/10.1007/s10032-014-0218-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10032-014-0218-7