Devanagari Character Classification Using Capsule Network

  • Jeel Sukhadiya
  • Yashi SubaEmail author
  • Mitchell D’silva
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 940)


Handwritten character recognition has been a widely explored topic in computer vision since its inception and exceptional works have been proposed in this area. Evolving over the primordial works in this domain, Deep Convolutional neural networks provide an epitome solution to classification of characters but still lack orientation and spatial relationships among entities. Convolutional neural networks fail in capturing the pose, orientation and view of the images due to the inefficiency of max pooling layer. These limitations are overcome by a novel approach: Capsule Networks. Capsule Networks are made up of layers of capsules representing the instantiation parameters of entities by using the dynamic routing and route by agreement algorithms.


Capsule Network Histogram oriented gradients Support vector machines Feed forward neural networks 


  1. 1.
    Gader, P.D., Keller, J.M., Krishnapuram, R., Chiang, J.-H., Mohamed, M.A.: Neural and fuzzy methods in handwriting recognitionGoogle Scholar
  2. 2.
    Stallkamp, J., Schlipsing, M., Salmen, J., Igel, C.: The German traffic sign recognition benchmark: a multi-class classification competitionGoogle Scholar
  3. 3.
    Hinton, G., Sabour, S., Frosst, N.: Matrix capsules with EM routing. Published as a Conference Paper at ICLR 2018 (2018)Google Scholar
  4. 4.
    Iamsa-at, S., Horata, P.: Handwritten character recognition using histograms of oriented gradient features in deep learning of artificial neural network, Conference Paper, December 2013Google Scholar
  5. 5.
  6. 6.
    Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego USA, pp. 886–893, June 2005Google Scholar
  7. 7.
  8. 8.
  9. 9.
    Svozil, D., Kvasnicka, V., Pospichal, J.: Introduction to multi-layer feed-forward neural networks. Chemom. Intell. Lab. Syst. 39, 43–62 (1997)CrossRefGoogle Scholar
  10. 10.
    Kamblea, P.M., Hegadi, R.S.: Handwritten Marathi character recognition using R-HOG feature. In: International Conference on Advanced Computing Technologies and Applications (ICACTA 2015) (2015)CrossRefGoogle Scholar
  11. 11.
    Acharya, S., Pant, A.K., Gyawali, P.K.: Deep learning based large scale handwritten Devanagari character recognition. In: 9th International Conference on Software, Knowledge, Information Management and Applications (SKIMA) (2015)Google Scholar
  12. 12.
  13. 13.
    Sabour, S., Frosst, N., Hinton, G.E.: Dynamic routing between capsules. In: 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA (2017)Google Scholar
  14. 14.
    Sharma, N., Pal, U., Kimura, F., Pal, S.: Recognition of off-line handwritten Devnagari characters using quadratic classifier. In: ICVGIP. LNCS, vol. 4338, pp. 805–816. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  15. 15.
    Basu, S., Das, N., Sarkar, R., Kundu, M., Nasipuri, M., Basu, D.K.: A novel framework for automatic sorting of postal documents with multi-script address blocks. Pattern Recognit. 43, 3507–3521 (2010)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  • Jeel Sukhadiya
    • 1
  • Yashi Suba
    • 1
    Email author
  • Mitchell D’silva
    • 1
  1. 1.Department of Information TechnologyDwarkadas J. Sanghvi College of EngineeringMumbaiIndia

Personalised recommendations