Skip to main content
Log in

A novel normal to tangent line (NTL) algorithm for scale invariant feature extraction for Urdu OCR

  • Original Paper
  • Published:
International Journal on Document Analysis and Recognition (IJDAR) Aims and scope Submit manuscript

Abstract

The font invariant recognition of Urdu optical characters is a difficult task due to the nature of Nastalique script. Urdu Nastalique is a complex script as it is excessively cursive and contains characters which are overlapping. Characters also change shape along with change in context. The identification of starting position of same character in different contexts further increases complexity. Hence, an optical character recognition (OCR) system, which is trained to recognize characters of a particular font size, may not show the same level of accuracy if font size varies. While considering this complexity the current research has focused on discovering such a feature set which may provide sufficient information for scale invariant Urdu optical character recognition. For this task, calligraphic properties of Urdu Nastalique, the thickness of ligature, the direction of movement of calligraphic pen and global geometric features (height and weight) are used as feature set. The feature of thickness is extracted using two novel algorithms, i.e. “Normal to Tangent Line Algorithm (NTL)” and “Angle to Tangent Line Algorithm (ATL)”. These features are fed to three different models, i.e. correlation, C4.5 and feedforward artificial neural network, and the performance of these models is also compared with SIFT (Scale Invariant Features Transformation). For training and testing, both real and fabricated data sets are employed. The new benchmark dataset of extracted features named Urdu OCR—Scale Invariant Feature Vectors (SIFVs), is developed and released at Kaggle. The newly developed SIFVs dataset, when used to train Correlation, C4.5 and ANN-based models, outperformed SIFT descriptors and yielded 94.56%, 90.54% and 94.65% accuracy, respectively, while SIFT descriptors achieved only 75.45% accuracy on average.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17

Similar content being viewed by others

Availability of data and material

Datasets are available at the given references.

References

  1. Ball, G.R., Srihari, S.N., Srinivasan, H.: Segmentation-based and segmentation-free methods for spotting handwritten Arabic words. In: Proceedings of Tenth International Workshop on Frontiers in Handwriting Recognition (2006).

  2. Avrithis, Y., Delopoulos, A., Kollias, S.: An efficient scheme for invariant optical character recognition using triple correlations. In: International Conference on Computer Applications to Engineering Systems, Nicosia, Cyprus (1993).

  3. Martin, A., Tosunoglu, S.: Image processing techniques for machine vision. In: Conference on Recent Advances in Robotics, Miami, Florida, (1986).

  4. Lowe, D.G.: Object recognition from local scale-invariant features. In: Proceedings of International Conference on Computer, Vision 2, pp. 1150–1157 (1999).

  5. Naseer, A., Zafar, K.: Meta features-based scale invariant OCR decision making using LSTM-RNN. Comput. Math. Org. Theory 25, 165–183 (2018)

    Article  Google Scholar 

  6. Rehman, B., Halim, Z., Ahmad, M.: ASCII based GUI system for arabic scripted languages: a case of Urdu. Int. Arab J. Inf. Technol. 11(4), 329–337 (2014)

    Google Scholar 

  7. Asad, M., Halim, Z., Waqas, M., Tu, S.: An In-ad contents-based viewability prediction framework using artificial intelligence for web Ads. Artif. Intell. Rev. (2021). https://doi.org/10.1007/s10462-021-10013-3

    Article  Google Scholar 

  8. Halim, Z., Rehan, M.: On identification of driving-induced stress using electroencephalogram signals: a framework based on wearable safety-critical scheme and machine learning. Inf. Fusion 53, 66–79 (2020)

    Article  Google Scholar 

  9. Sohail, A.S., Haque, S., Pathan, M.K.: Nastaliq optical character recognition. In: Proceedings of 46th Annual Southeast Regional Conference, Auburn, Alabama (2008)

  10. Kour, H., Gondhi, N.K.: Machine Learning approaches for Nastaliq style Urdu handwritten recognition: a survey. In: 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), 2020, pp. 50–54 (2020). https://doi.org/10.1109/ICACCS48705.2020.9074294.

  11. Mannas, A., Babri, H.A., Saeed, M.: Circular sector based feature extraction for offline binary symbol recognition. In: Proceedings of International Conference on Intelligence and Information Technology (ICIIT), Lahore, Pakistan (2010).

  12. Bose, C.B., Kuo, S.: Connected and degraded text recognition using hidden Markov model. Pattern Matching 27(10), 1345–1363 (1994)

    Google Scholar 

  13. Govindaraju, V., Srihari, S.N.: OCR in a hierarchical feature space. IEEE Trans. Pattern Anal. Mach. 22, 400–407 (2000)

    Article  Google Scholar 

  14. Sattar, S.A.: A technique for the design and implementation of an OCR for printed nastalique text. Ph.D. thesis, NED University of Engineering and Technology, Karachi, Pakistan (2009)

  15. Ahmad, Z, Orakzai, J.K., Shamsher, I., Adnan, A.: Urdu nastaleeq optical character recognition. In: Proceedings of World Academy of Science, Engineering And Technology, vol. 26 (2007).

  16. Mushtaq, F., Misgar, M.M., Kumar, M., Khurana, S.S.: UrduDeepNet: offline handwritten Urdu character recognition using deep neural network. Neural Comput. Appl. (2021). https://doi.org/10.1007/s00521-021-06144-x

    Article  Google Scholar 

  17. Lu, Z., Bazzi, I., Kornai, A., Makhoul, J.: A robust, language independent OCR system. In: 27th Advances in Computer-Assisted Recognition SPIE (1999).

  18. Pal, U., Sarkar, A.: Recognition of printed Urdu script. In: Proceedings of Seventh International Conference on Document Analysis and Recognition, pp. 1183–1187 (2003).

  19. KO, M.A., Poruran, S.: OCR-nets: variants of pre-trained CNN for Urdu handwritten character recognition via transfer learning. Procedia Comput. Sci. 171, 2294–2301 (2020)

    Article  Google Scholar 

  20. Arafat, S.Y., Ashraf, N., Iqbal, M.J., Ahmad, I., Khan, S., Rodrigues, J.J.: Urdu signboard detection and recognition using deep learning. Multimedia Tools Appl (2021). https://doi.org/10.1007/s11042-020-10175-2

    Article  Google Scholar 

  21. Naseer, A.: Urdu OCR—Scale Invariant Feature Vectors, Version 1 (2021). https://www.kaggle.com/asmanaseer/urdu-ocr-scale-invariant-feature-vectors.

  22. Amin, A.R., Khan, S.H., Khan, M.A.U.: Scale and rotation invariant recognition of cursive pashto script using SIFT features. In: 6th International Conference on Emerging Technologies (ICET), Islamabad, Pakistan (2010).

  23. Naseer, A., Hussain, S.: Supervised word sense disambiguation for Urdu using Bayesian classification. Center for Research in Urdu Language Processing, Lahore, Pakistan (2009).

  24. Obaid, A.M.: A new pattern matching approach to the recognition of printed Arabic. In: Workshop Computational Approaches to Semitic Languages, Budapest, Hungry (1998).

  25. Inpage Urdu Software. http://www.inpage.com.

  26. Naseer, A., Zafar, K.: Comparative analysis of raw images and meta feature based Urdu OCR using CNN and LSTM. Int. J. Adv. Comput. Sci. Appl. (IJACSA) 9(1), 419–424 (2018)

    Google Scholar 

  27. Naseer, A., Zafar, K.: Scale invariant features for Urdu optical character recognition, Ph.D. Thesis, National University of Computer and Emerging Sciences (NUCES-FAST) (2019).

  28. Rehmam, B., Halim, Z., Abbas, G., Muhammad, T.: Artificial neural network-based speech recognition using dwt analysis applied on isolated words from oriental languages. Malays. J. Comput. Sci. 28(3), 242–262 (2015)

    Article  Google Scholar 

Download references

Funding

Not applicable.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Kashif Zafar.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Naseer, A., Hussain, S., Zafar, K. et al. A novel normal to tangent line (NTL) algorithm for scale invariant feature extraction for Urdu OCR. IJDAR 25, 51–66 (2022). https://doi.org/10.1007/s10032-021-00389-x

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10032-021-00389-x

Keywords

Navigation