Handwritten Off-line Kannada Character/Word Recognition Using Hidden Markov Model

  • G. S. VeenaEmail author
  • T. N. R. Kumar
  • A. Sushma
Conference paper
Part of the Lecture Notes in Networks and Systems book series (LNNS, volume 14)


Digitization of handwritten documents is a challenging task in the area of character recognition, because of the variations in font style and font size in writing character. An effort is made to design a classifier which can handle the different variation in font size, font style, overlapping of characters and partially visible written characters. An effort is made by considering geometrical structure of the character. In this work two important components of feature extraction is used, one is gradient direction matrix and other another is aspect ratio. Each and every character image is subjected to preprocessing steps, further characters images are subjected to feature extraction process. The gradient based method used for feature extraction results in feature vectors which given as input to Hidden Markov Model training (HMM), and the test samples are tested against the trained models and results are analyzed. It is evident from the obtained result that the recognition rate on an average is around 66%.


Gradient based feature extraction Aspect ratio direction matrix Binarization Hidden markov model 


  1. 1.
    Sharma N, Pal U, Kimura F (2006) Recognition of handwritten Kannada numerals. In: Proceedings of IEEE-ICIT 2006, pp 133–136. doi: 10.1109/ICIT.2006.77
  2. 2.
    Arica N, Yarman-Vural F (2001) An overview of character recognition focused offline handwriting. 31(2):216–233. doi: 10.1109/5326.941845
  3. 3.
    Kumar S (2005) Vikas-online character recognition. In: Third international ICITA’05 2005, vol 1, pp 698–703. doi: 10.1109/ICITA.2005.197
  4. 4.
    Jagadeesh Kumar R, Prabhakar R, Suresh RM (2008) Off-line cursive handwritten Tamil characters recognition. In: International conferences on security technology, pp 159–164. doi: 10.1109/SecTech.2008.33
  5. 5.
    Shridhar M, Badreldin A (1985) High accuracy syntactic recognition algorithm for handwritten numerals. IEEE Trans Syst Man Cybern SMC15 (1):152–158. doi: 10.1109/TSMC.1985.6313404
  6. 6.
    Akshay Parvatikar AGS, Veena GS (2015) Use of hidden markov’s model for handwritten Kannada character recognition. Int J Comput Sci Inform Technol 6(3):2959–2960GGoogle Scholar
  7. 7.
    Roy PP, Dey P, Roy S, Pal U, Kimura F (2014) A novel approach of Bangla handwritten text recognition using HMM. In: 14th international conference on frontiers in handwriting recognition, Heraklion, pp 661–666. doi: 10.1109/ICFHR.2014.116
  8. 8.
    Huang BQ, Zhang YB, Kechadi MT (2007) Preprocessing techniques for online handwriting recognition. In: Seventh international conference on intelligent systems design and applications, Rio de Janeiro, pp 793–800. doi: 10.1109/ISDA.2007.31
  9. 9.
    Mathew CJ, Shinde RC, Patil CY (2015) Segmentation techniques for handwritten script recognition system. In: International conference on circuit, power and computing technologies, 2015, Nagercoil, pp 1–7. doi: 10.1109/ICCPCT.2015.7159397
  10. 10.
    Wu X, Tang Y, Bu W (2014) Offline text-independent writer identification base on scale invariant feature transform. IEEE Trans Inf Forensics Secur 9(3):526–536Google Scholar
  11. 11.
    Subashini A, Kodikara ND (2011) A novel SIFT-based codebook generation for handwritten Tamil character recognition. In: 2011 6th international conference on industrial and information systems, Kandy, pp 261–264. doi: 10.1109/ICIINFS.2011.6038077
  12. 12.
    Ding J, Li G, Wen C, Chua CS (2014) Min-max discriminant analysis based on gradient method for feature extraction. In: 2014 13th international conference on control automation robotics & vision, Singapore, 2014, pp 129–134. doi: 10.1109/ICARCV.2014.7064292
  13. 13.
    Ameur H, Helali A, Nasri M, Maaref H, Youssef A (2014) Improved feature extraction method based on histogram of oriented gradients for pedestrian detection. In: Global summit on computer information technology 2014, Sousse, pp 1–5. doi: 10.1109/GSCIT.2014.6970120
  14. 14.
    El Yacoubi MA (2004) Offline handwritten word recognition using HMM. citseerGoogle Scholar
  15. 15.
    Primekumar KP, Idiculla SM (2013) On-line Malayalam handwritten character recognition using HMM and SVM. In: 2013 international conference on signal processing image processing & pattern recognition, Coimbatore, pp 322–326. doi: 10.1109/ICSIPR.2013.6497991
  16. 16.
    Zhang J, Zhou W, Xie C, Pu J, Li H (2016) Chinese sign language recognition with adaptive HMM. In: 2016 IEEE international conference on multimedia and expo, Seattle, WA, USA, pp 1–6. doi: 10.1109/ICME.2016.7552950
  17. 17.
    El Moubtahij H, Halli A, Satori K (2016) Recognition of off-line arabic handwriting words using HMM toolkit. In: 2016 13th international conference on computer graphics, imaging and visualization, Beni Mellal, pp 167–171. doi: 10.1109/CGiV.2016.40
  18. 18.
    Maqqor A, Halli A, Satori K, Tairi H (2014) Using HMM toolkit (HTK) for recognition of arabic manuscripts characters. In: International conference on multimedia computing and systems, 2014 , Marrakech, pp 475–479. doi: 10.1109/ICMCS.2014.6911316
  19. 19. Elissa, Title of paper if known, (unpublished)

Copyright information

© Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  1. 1.Department of Computer Science and EngineeringM.S.R.I.TBengaluruIndia

Personalised recommendations